Presented By O’Reilly and Intel AI
Put AI to work
8-9 Oct 2018: Training
9-11 Oct 2018: Tutorials & Conference
London, UK

Speakers

Hear from innovative programmers, talented managers, and senior developers who are doing amazing things with artificial intelligence. More speakers will be announced; please check back for updates.

Filter

Search Speakers

Zack Akil is a developer advocate for Cloud Machine Learning at Google. When he’s not teaching machine learning at Google, he likes to teach machine learning at his hands-on data science meetup, Central London Data Science Project Nights. Although he works in the cloud, most of his hobby projects look at different ways you can embed machine learning into low-power devices like Raspberry Pis and Arduinos. He also likes to have a bit of banter with his mixed tag rugby teams.

Presentations

Pragmatic ML development with scikit-learn and TensorFlow using Google ML Engine (sponsored by Google) Session

Zack Akil shares pragmatic techniques and useful tools that can help you avoid common pitfalls when building ML, including tools for notebook collaboration and version control that will help prevent you and your teammates from stepping on each others' toes as well as an iterative ML model development approach that will prevent your project from stagnating.

Sabih Ali directs strategy at After the Flood, a leading data design consultancy designing platforms and solutions for large corporations to take to market. Previously, Sabih was head of brand, lead web designer, and digital analytics manager at leading global software firm Symplectic. His passion has always been the intersection of business and technology, bringing together his background in science and engineering with a focus on user-centered design thinking. He holds a degree in electronic engineering from UCL and a master’s degree in nanotechnology enterprise from the University of Cambridge.

Presentations

Executive Briefing: Why designing for trust matters Session

Max Gadney and Sabih Ali explore some of the ways designers and product teams are designing systems with transparency, trust, and privacy in mind.

Alasdair Allan is a scientist and researcher who has authored more than 80 peer-reviewed papers and eight books and has been involved with several standards bodies. Originally an astrophysicist, Alasdair now works as a consultant and journalist, focusing on open hardware, machine learning, big data, and emerging technologies, with expertise in electronics, especially wireless devices and distributed sensor networks, mobile computing, and the internet of things. He runs a small consulting company and has written for Make: magazine, Motherboard/VICE, Hackaday, Hackster.io, and the O’Reilly Radar. In the past, he has mesh-networked the Moscone Center, caused a US Senate hearing, and contributed to the detection of what was at the time the most distant object yet discovered.

Presentations

Do-it-yourself artificial intelligence Session

Google's AIY Projects kits bring Google's machine learning algorithms to developers with limited experience in the field, allowing them to prototype machine learning applications and smart hardware more easily. Alasdair Allan walks you through setting up and building the kits and demonstrates how to use the kits' Python SDK for machine learning both in the cloud and locally on a Raspberry Pi.

Johnnie Ball is chief data scientist at Fluidly, a company that applies machine learning to transaction data from cloud accounting packages and bank accounts to automate the forecasting, optimization, and protection of cashflow. Johnnie spent his early career as an interest rates trader at an major investment bank before deciding to return to academia to study computational statistics and machine learning at UCL. Over his subsequent career, Johnnie has been a data scientist at a hedge fund and an energy company and a cohort member at deep tech startup accelerator Entrepreneur First.

Presentations

AI in business forecasting: Lessons from building an intelligent cashflow engine Session

Cashflow is responsible for 80–90% of UK SME failure. Fluidly uses the wealth of financial data available through APIs to instantly predict cashflow. Johnnie Ball details how the company built an automated cashflow engine, explores the challenges faced in applying AI to financial data, and explains how machine learning can redefine how we think about established approaches to modeling.

Jonathan Ballon is vice president of the Internet of Things Group at Intel, where he is responsible for a global team chartered with driving and accelerating innovation and growth in various market segments. His team is a pioneer in artificial intelligence and deep learning applications with computer vision capabilities, supported by software tools and a robust ecosystem. Jonathan also leads the IoT channel routes to market and is responsible for Intel’s China engineering teams and business. He is passionate about the application of technology toward the environment, human productivity, safety, education, health, and longevity.

Presentations

AI in production: The droids you’re looking for Keynote

Artificial intelligence in the future, at least represented in science fiction, can learn, interpret, and take action based on data analysis. AI in production is the present, a present that feels decidedly futuristic. Jonathan Ballon explains why Intel’s leading portfolio of AI and computer vision edge technology will drive advances that improve how we work and live.

David Barber is reader in information processing in the Department of Computer Science at UCL, where he develops novel machine learning algorithms. David is also a cofounder of the NLP company reinfer.io.

Presentations

AI and the challenges that remain

While great strides have been made in perceptual AI (for example, in speech recognition), there's been relatively modest progress in reasoning AI—systems that can interact with us in natural ways and understand the objects in our environment. David Barber explains why general AI will be out of reach until we address how to endow machines with knowledge of our environment.

Louis Barson is Deputy Director of the Future Sectors team in the Department for Business, Energy, and Industrial Strategy. The Future Sectors team lead the Industrial Strategy approach to new and emerging sectors, such as Robotics, Quantum Technologies and Drones, as well as ministerial reviews into clusters of disruption in ‘transforming sectors’ such as EdTech, InsurTech and GovTech. The team led work on the Artificial Intelligence policy framework in partnership with DCMS leading up to the Grand Challenge, Sector Deal, and establishment of the new Office for AI.

Prior to this role, Louis was Head of Technology for Aerospace, Marine and Defence, responsible for the £3.9bn Aerospace R&D budget, and Deputy Director for Science and Innovation in UK Trade and Investment (UKTI), where he was responsible for leading the UKTI Innovation Gateway – a ‘front door’ to the UK innovation system for large international investors and corporates.

Prior to joining the Civil Service, Louis spent four years in Japan studying Japanese language to fluency, and carrying out post-graduate research into Japanese Philosophy and non-classical Mathematical Logic. He graduated from Kings College London in 2005 with a first-class honours degree in Philosophy. In his spare time, he is studying Physics with the Open University and enjoys recreational flying as a private pilot.

Presentations

Fireside chat with Marc Warner and Louis Barson Keynote

Fireside chat with Marc Warner and Louis Barson

Denis Batalov is a principal solutions architect at AWS EMEA, helping customers adopt AI and ML technologies. A 13-year Amazon veteran, Denis has worked on such exciting projects as Search Inside the Book, Amazon Mobile apps, and Kindle Direct Publishing. He is a frequent public speaker. Denis holds a PhD in machine learning. You can follow him on Twitter at @dbatalov.

Presentations

Building deep learning applications with Amazon SageMaker Tutorial

Join Denis Batalov for an overview of the Amazon SageMaker machine learning platform. Denis walks you through setting up an Amazon SageMaker notebook (a hosted Jupyter Notebook server), using a built-in SageMaker deep learning algorithm, and building your own neural network architecture using SageMaker's prebuilt TensorFlow containers.

Rachel Bellamy is a principal research scientist and manages the Human-AI Collaboration Group at the IBM T.J. Watson Research Center in Yorktown Heights, New York, where she leads an interdisciplinary team of human-computer interaction experts, user experience designers, and user experience engineers. Previously, she worked in the Advanced Technology Group at Apple, where she conducted research on collaborative learning and led an interdisciplinary team that worked with the San Francisco Exploratorium and schools to pioneer the design, implementation, and use of media-rich collaborative learning experiences for K–12 students. She holds many patents and has published more than 70 research papers. Rachel holds a PhD in cognitive psychology from the University of Cambridge and a BS in psychology with mathematics and computer science from the University of London.

Presentations

Designing user interfaces for AI for unbiased decision making Session

Data bias is not only an AI problem; it's also a UI problem. Non-AI experts use custom application interfaces to help them make decisions based on predictions from machine learning models. These application interfaces need to be designed so that the decisions made are unbiased. Rachel Bellamy and Casey Dugan explain how to represent model predictions so that people can recognize if they are fair.

Chris Boyd heads up The Wall Street Journal’s digital advertising and membership product management and engineering efforts.  Chris spent most of the last 10 years leading the engineering organization behind the wsj.com website as consumer technology and news consumption habits have gone through tremendous change.  The last few years Chris led efforts to create an award winning customer and data driven paywall and introduced machine learning models to power customer engagement that has lead to The Wall Street Journal’s highest levels of membership ever.  Prior to his work at the Journal, Chris worked in consulting as a technical architect of mission critical enterprise applications across a variety of industries.

Presentations

The WSJ dynamic paywall Session

Chris Boyd and John Wiley explain how the Wall Street Journal uses machine learning and a proprietary algorithm to predict the likelihood for someone subscribing, which in turn dictates the paywall experience that customer receives.

Paul Brasnett is a principal research engineer leading a team responsible for imaging, image processing, and computer vision algorithms within the PowerVR Division of Imagination Technologies. Previously, he was a senior research engineer at Mitsubishi Electric, where he was involved in image and video processing and participated in MPEG standardization work. Paul holds a PhD and MEng from the University of Bristol, UK.

Presentations

Enabling traditional vision on specialized deep learning hardware Session

In recent years, we’ve seen a shift from traditional vision algorithms to deep neural network algorithms. While many companies expect to move to deep learning for some or all of their algorithms, they may have a significant investment in classical vision. Paul Brasnett explains how to express and adapt a classical vision algorithm to become a trainable DNN.

Mikio Braun is principal engineer for search at Zalando, one of Europe’s biggest fashion platforms. He worked in research for a number of years before becoming interested in putting research results to good use in the industry. Mikio holds a PhD in machine learning.

Presentations

Architecting AI applications

Mikio Braun looks back on the past 20 years of machine learning research to explore aspects of artificial intelligence. He then turns to current examples like autonomous cars and chatbots, putting together a mental model for a reference architecture for artificial intelligence systems.

Cormac Brick is director of machine intelligence in the Movidius Group at Intel Corporation, where he builds new foundational algorithms for computer vision and machine intelligence to enhance the Myriad VPU product family. Cormac contributes to internal architecture and helps customers build products using the very latest techniques in deep learning and embedded vision through a set of advanced applications and libraries. He has worked with Movidius since its early days and has contributed heavily to the design of the ISA and the hardware systems as well as computer vision software development and tools. Cormac holds a BEng in electronic engineering from University College Cork.

Presentations

Portability and performance in embedded deep learning: Can we have both? Session

Recent research has shown that training for quantization can lead to large gains in energy efficiency, and embedded runtime packages like TensorFlow Lite and Caffe2Go offer portability over a number of platforms. Cormac Brick asks, Why can't we have both performance and portability? Cormac explores industry challenges and details the progress needed to close the portability-performance gap.

Gary Brown leads product marketing and strategic partnerships for Movidius, a division of Intel’s new Technology Group, where he draws on his background in embedded DSP and his passion for computer vision and deep learning technology to navigate Movidius into this new era of machine intelligence. He holds an MS in electrical engineering from Stanford University.

Presentations

The evolution of AI at the network edge: How silicon is paving a path for IoT innovation (sponsored by Intel AI) Session

Gary Brown explains how the use of AI in the IoT is leading to fascinating growth in various applications from industrial and medical to smart transportation and retail. Gary discusses Intel’s unique vantage point of the platforms paving the way for interesting new AI experiences. Along the way, he shares Intel’s latest IoT innovations.

Patrick Buehler is a senior data scientist at Microsoft Boston and has been in the field for over 10 years. His main interests are machine learning and computer vision. He holds a PhD from the VGG group at Oxford.

Presentations

Building end-to-end computer vision solutions from pretrained deep learning models Session

Dramatic progress has been made in computer vision. Deep neural networks (DNNs) trained on millions of images can recognize thousands of different objects, and they can be customized to new use cases. Vanja Paunic and Patrick Buehler outline simple methods and tools that enable users to easily and quickly adapt Microsoft's state-of-the-art DNNs for use in their own computer vision solutions.

Peter Cahill is the founder and CEO of Voysis. He has over 15 years’ experience in speech technology and neural network R&D. Previously, Peter was part of a group of scientists that attracted a total of $117M funding for ADAPT (formerly CNGL), a dynamic research center that combines leading academic researchers with key industry partners to produce groundbreaking digital content innovations. Peter is an active member of the speech research community; he chairs SynSIG, the global speech synthesis special interest group, serves as a reviewer for all leading journals and conferences in his field. He holds a PhD from University College Dublin, where he was also a faculty member.

Presentations

The second generation of voice interfaces Session

Peter Cahill explains why Wavenet will be the next generation of recognition, synthesis, and voice-activity detection.

Yishay Carmiel is the founder of IntelligentWire, a company that develops and implements industry-leading deep learning and AI technologies for automatic speech recognition (ASR), natural language processing (NLP), and advanced voice data extraction, and the head of Spoken Labs, the strategic artificial intelligence and machine learning research arm of Spoken Communications. Yishay and his teams are currently working on bleeding-edge innovations that make the real-time customer experience a reality—at scale. Yishay has nearly 20 years’ experience as an algorithm scientist and technology leader building large-scale machine learning algorithms and serving as a deep learning expert.

Presentations

How to build privacy and security into deep learning models Session

In recent years, there's been a quantum leap in the performance of AI, as deep learning made its mark in areas from speech recognition to machine translation and computer vision. However, as artificial intelligence becomes increasingly popular, data privacy issues also gain traction. Yishay Carmiel reviews these issues and explains how they impact the future of deep learning development.

Philip Carnelley is AVP for IDC’s European Software Group, where he leads IDC’s European research on enterprise applications and analytics, with a particular focus on big data and cognitive/AI software and solutions and their implications for digital business transformation. He heads IDC’s multidisciplinary European Cognitive/AI Research Practice, and he also oversees IDC’s European Software Tracker analyst team. He is also tasked with writing and presenting on broader enterprise software trends and the impact of emerging software technologies in the European region. Philip has over 20 years’ experience of working in the IT industry as an analyst, consultant, and applications developer in a variety of industries around Europe. He is a regular keynote speaker and panelist at industry events.

Presentations

Executive Briefing: Supporting digital business transformation through AI everywhere Session

AI is a key innovation accelerator for digital business transformation. To help you with your strategic roadmap, Philip Carnelley shares IDC's research into the AI market across hundreds of European organizations and explains why organizations should establish a digital platform based on big data, AI, and cloud technologies, with an intelligent core, as part of their transformation strategy.

Gaurav Chakravorty is the cofounder and head of strategy development at qplum, a digital wealth management firm that specializes in family offices and HNWI. Qplum also works with endowment funds and pension funds to provide outsourced asset management services that integrate with the company’s in-house technology. Gaurav has been one of the early pioneers in machine learning-based high-frequency trading. He built the most profitable algo-trading group at Tower Research and was the youngest partner in the firm. Gaurav has been a guest speaker on a few popular podcasts.

Presentations

The use of recommender systems in the chief investment office: A case study Session

Gaurav Chakravorty explains how recommender systems can be utilized for investment management and details how AI and deep learning are used in trading today.

Pin-Yu Chen is a research staff member in the AI Foundations Learning Group at the IBM Thomas J. Watson Research Center in Yorktown Heights, NY. His recent research focuses on adversarial machine learning and robustness analysis of neural networks; he’s also interested in graph and network data analytics and their applications to data mining, machine learning, signal processing, and cybersecurity. Pin-Yu received the NIPS 2017 Best Reviewer Award and the IEEE GLOBECOM 2010 GOLD Best Paper Award as well as several travel grants, including IEEE ICASSP 2014 (NSF), IEEE ICASSP 2015 (SPS), IEEE Security and Privacy Symposium, NSF Graph Signal Processing Workshop 2016, and ACM KDD 2016. He is a member of the Tau Beta Pi Honor Society and the Phi Kappa Phi Honor Society and was the recipient of the Chia-Lun Lo Fellowship from the University of Michigan Ann Arbor. Pin-Yu holds a BS in electrical engineering and computer science (undergraduate honors program) from National Chiao Tung University, Taiwan, an MS in communication engineering from National Taiwan University, Taiwan, and an MA in statistics and a PhD in electrical engineering and computer science, both from the University of Michigan, Ann Arbor.

Presentations

How CLEVER is your neural network? Robustness evaluation against adversarial examples Session

Neural networks are particularly vulnerable to adversarial inputs. Carefully designed perturbations can lead a well-trained model to misbehave, raising new concerns about safety-critical and security-critical applications. Pin-Yu Chen offers an overview of CLEVER, a comprehensive robustness measure that can be used to assess the robustness of any neural network classifiers.

Roger Chen is cofounder and CEO of Computable Labs and program chair for the O’Reilly Artificial Intelligence Conference. Previously, he was a principal at O’Reilly AlphaTech Ventures (OATV), where he invested in and worked with early-stage startups primarily in the realms of data, machine learning, and robotics. Roger has a deep and hands-on history with technology. Before startups and venture capital, he was an engineer at Oracle, EMC, and Vicor. He also developed novel nanoscale and quantum optics technology as a PhD researcher at UC Berkeley. Roger holds a BS from Boston University and a PhD from UC Berkeley, both in electrical engineering.

Presentations

Closing remarks Keynote

Program cochairs Ben Lorica and Roger Chen close the second day of keynotes.

Decentralized data markets for training AI models

Blockchain technologies offer new internet primitives for creating open and online data marketplaces. Roger Chen explores how data markets can be constructed and how they offer a shared resource on the internet for AI-based research, discovery, and development.

The state of automation technologies Keynote

What technologies are ready for adoption, and how should companies and organizations evaluate automation technologies? Ben Lorica and Roger Chen highlight recent trends in data, compute, and machine learning.

Thursday opening remarks Keynote

Program cochairs Ben Lorica and Roger Chen open the second day of keynotes.

Wednesday opening remarks Keynote

Program cochairs Ben Lorica and Roger Chen open the first day of keynotes.

Yijing Chen is a senior data scientist in the Cloud AI Group at Microsoft, where she works with external customers in areas such as energy demand forecast, user mobile behavioral analysis, retail demand forecast, energy theft detection, product pricing, and medical claim denial prediction as well as on other projects using various machine learning methods. Yijing holds an MA in statistics from Harvard University.

Presentations

Recurrent neural networks for time series forecasting Tutorial

Buisnesses use forecasting to make better decisions and allocate resources more effectively. Recurrent neural networks (RNNs) have achieved a lot of success in text, speech, and video analysis but are less used for time series forecasting. Join Yijing Chen, Dmitry Pechyoni, Angus Taylor, and Vanja Paunic to learn how to apply RNNs to time series forecasting.

Christopher Cho is a product manager and cloud program manager at Google, where he helps customers solve machine learning and infrastructure problems, and is one of the product managers in Kubeflow team. Previously, Chris was research program manager at DeepMind, working on cutting-edge ML research. His background is in enterprise business consulting. Chris is currently working toward his MSCS at Georgia Tech. He holds a BS in mechanical engineering from the University of Illinois Urbana-Champaign.

Presentations

Machine learning at scale with Kubernetes Session

Christopher Cho details how to leverage Kubernetes and the mighty Kubernetes APIs to build a complete deep learning pipeline, from data ingestion and aggregation to preprocessing and ML training to serving. Along the way, Christopher covers Kubeflow, a Google open source solution for managing machine learning with TensorFlow in a portable, scalable manner.

Michael Chui is a San Francisco-based partner in the McKinsey Global Institute, where he directs research on the impact of disruptive technologies, such as big data, social media, and the internet of things, on business and the economy. Previously, as a McKinsey consultant, Michael served clients in the high-tech, media, and telecom industries on multiple topics. Prior to joining McKinsey, he was the first chief information officer of the City of Bloomington, Indiana, and was the founder and executive director of HoosierNet, a regional internet service provider. Michael is a frequent speaker at major global conferences and his research has been cited in leading publications around the world. He holds a BS in symbolic systems from Stanford University and a PhD in computer science and cognitive science and an MS in computer science, both from Indiana University.

Presentations

Notes from the frontier: Making AI work Keynote

Drawing on the McKinsey Global Institute's groundbreaking research, Michael Chui explores commonly asked questions relating to AI and its impact on work. Michael also previews new research showing that despite the rapid pace of AI adoption, much foundational work in enterprises remains to be done to capture value at scale.

Valeriu Codreanu is the PI of the Intel Parallel Computing Center at SURFsara, focusing on optimizing deep learning techniques using the Intel ecosystem as well as extending the use of these techniques to other scientific domains. Previously, he was an HPC consultant at SURFsara, focusing on machine learning and a postdoctoral researcher at both Eindhoven and Groningen Universities, working on GPU computing, computer vision, and embedded systems in the scope of several EU-funded projects. Valeriu holds an MSc in electrical engineering and a PhD in computer architecture from the Polytechnic University of Bucharest.

Presentations

Efficient neural network training on Intel Xeon-based supercomputers Session

SURFSara and Intel collaborated as part of the Intel Parallel Computing Center initiative to advance the state of large-scale neural network training on Intel Xeon CPU-based servers. Ananth Sankar, Valeriu Codreanu, Damian Podareanu, and Steve Smith share insights on several best-known methods for neural network training and present results from tests performed on Stanford's CheXNet project.

Ira Cohen is a cofounder and chief data scientist at Anodot, where he is responsible for developing and inventing the company’s real-time multivariate anomaly detection algorithms that work with millions of time series signals. He holds a PhD in machine learning from the University of Illinois at Urbana-Champaign and has over 12 years of industry experience.

Presentations

Algorithms gone wild: Applying machine learning for insights into machine learning algorithms Session

With the more applications of machine learning-based applications, the complex algorithms that automate behaviors can get out of control. Ira Cohen explains how to catch problems and glitches early on by using machine learning algorithms to monitor these algorithms for anomalous behavior.

James Crawford is the founder and CEO of Orbital Insight, where he leads the company’s efforts to leverage artificial intelligence to create geospatial analytics for an interconnected world. Previously, Jimi was the SVP of science and engineering at the Climate Corporation; CTO and software architect at Moon Express, whose goal was putting the first commercial robot on the moon; and engineering director for Google Book Search, where he was in charge of Google’s project to scan, index, and make searchable all the world’s books. He has also spent time at the NASA Ames Research Center, leading autonomy and robotics projects. He has authored over 15 peer-reviewed publications, resulting in five patents. Jimi holds both a PhD and master’s degree in computer science from the University of Texas at Austin and a BA in math and computer science from Rice University.

Presentations

How AI is taking geospatial data from alternative to mainstream in finance Session

By some estimates, soon it will require eight million people doing nothing but looking at satellite imagery 24/7 in order to ensure every photo taken on a daily basis is viewed. James Crawford explains how artificial intelligence solves this problem of scale, allowing us to accurately analyze reams of satellite imagery and detect patterns of socioeconomic change in a timely fashion.

Quentin de Laroussilhe is a Zurich-based software engineer on the applied machine intelligence team at Google AI, where he is helping various product areas deploy machine learning solutions for their use cases. He focuses on research topics including transfer learning, black-box optimization, and reinforcement learning. Previously, Quentin worked in image and voice search and on the Google Assistant.

Presentations

Machine learning in production with TensorFlow Extended (TFX) (sponsored by Google) Session

As machine learning evolves from experimentation to serving production workloads, so does the need to effectively manage the end-to-end training and serving workflow. Kenny Song and Quentin de Laroussilhe offer an overview of TensorFlow Extended, the end-to-end machine learning platform for TensorFlow that powers products across all of Google.

Benoit Dherin is a machine learning solutions engineer at Google’s Advanced Solutions Lab, where he teaches machine learning to Google clients and helps them implement machine learning solutions in the Google Cloud. Previously, Benoit held roles in data science, machine learning, and software engineering at various companies and startups in Silicon Valley as well as research and teaching positions at universities around the world. Benoit holds a PhD in mathematics from ETH Zurich. In his free time, he enjoys reading, traveling, and doing Bikram Yoga.

Presentations

Image classification models in TensorFlow Tutorial

Benoit Dherin explains how machine learning is applied to image classification, discusses evolving methods and challenges, and walks you through creating increasingly sophisticated image classification models using TensorFlow.

Serverless machine learning with TensorFlow on GCP Day (sponsored by Google Cloud) Event

Benoit Dherin leads an introduction to designing and building machine learning models on Google Cloud Platform. Through a combination of presentations, demos, and hand-ons labs, you'll learn machine learning (ML) and TensorFlow concepts and develop skills in developing, evaluating, and productionizing ML models.

Joshua V. Dillon is a software engineer for research and machine intelligence at Google. His research interests include approximate inference techniques for probabilistic models, uncertainty in machine learning, and designing probabilistic programming tools and languages. He holds a PhD in machine learning from the Georgia Institute of Technology. In his free time, Josh enjoys spending time with his family, cycling, and woodworking.

Presentations

Frontiers of TensorFlow: Mathematics and music (sponsored by Google) Session

Joshua Dillon and Wolff Dobson discuss core TensorFlow Probability (TFP) abstractions and demo some of TFP's modeling power and convenience. They also share some of the recent results from Project Magenta, a research project exploring the role of machine learning in the process of creating art and music.

Wolff Dobson is a developer programs engineer at Google specializing in machine learning and games. Previously, he worked as a game developer, where his projects included writing AI for the NBA 2K series and helping design the Wii Motion Plus. Wolff holds a PhD in artificial intelligence from Northwestern University.

Presentations

Frontiers of TensorFlow: Mathematics and music (sponsored by Google) Session

Joshua Dillon and Wolff Dobson discuss core TensorFlow Probability (TFP) abstractions and demo some of TFP's modeling power and convenience. They also share some of the recent results from Project Magenta, a research project exploring the role of machine learning in the process of creating art and music.

Rahul Dodhia is the director of data science in the Office of the President at Microsoft, currently leading a team of machine learning scientists and engineers in the Corporate, External, and Legal Affairs (CELA) Division. In collaboration with world-class researchers at Microsoft and its partners, his team is developing artificial intelligence applications for the legal industry. His experience in data and analytics spans 20 years, with emphases on statistics, machine learning, engineering for data science, and business applications. Previously, he worked at NASA’s Ames Research Center, Amazon, and Expedia and consulted for several startups. Rahul holds a PhD in mathematical psychology from Columbia University.

Presentations

Legal contract review by an artificial intelligence Session

Artificial intelligence is mature enough to make substantial contributions to the legal industry. Rahul Dodhia offers an overview of an AI assistant that can perform routine tasks such as contract review and checking compliance with regulations at higher accuracy rates than legal professionals.

Casey Dugan is the manager of AI experiences at IBM Research in Cambridge. She interned with the IBM Cambridge team twice before joining full time. In her time at IBM, she’s worked on projects like Malibu, Beehive (Social Blue), Blog Muse, TimeSquare/Timeflash, and Social Pulse. Her latest projects are the #selfiestation and the Meeting Room of the Future. She graduated from MIT.

Presentations

Designing user interfaces for AI for unbiased decision making Session

Data bias is not only an AI problem; it's also a UI problem. Non-AI experts use custom application interfaces to help them make decisions based on predictions from machine learning models. These application interfaces need to be designed so that the decisions made are unbiased. Rachel Bellamy and Casey Dugan explain how to represent model predictions so that people can recognize if they are fair.

Dmytro Dzhulgakov is an engineering manager and technical lead for AI infrastructure at Facebook, where he is currently leading the core development of PyTorch 1.0, an open source deep learning platform. Dmytro is one of the cocreators of ONNX, a joint initiative aimed at making AI development more interoperable. Previously, he built several generations of large-scale deep learning recommendation systems at Facebook that powered products from ads to the news feed. Dmytro holds an MS in applied mathematics. He had a successful career in programming competitions and was ranked in the Top 20 on Topcoder.

Presentations

Applied machine learning at Facebook: An infrastructure perspective Session

Machine learning sits at the core of many essential products and services at Facebook. Yangqing Jia and Dmytro Dzhulgakov offer an overview of the hardware and software infrastructure that supports machine learning at global scale.

PyTorch 1.0: Bringing research and production together Session

Dmytro Dzhulgakov explores PyTorch 1.0, from its start as a popular deep learning framework for flexible research to its evolution into an end-to-end platform for building and deploying AI models at production scale.

Daniel Ecer is an ASI Fellow and a data scientist at eLife Sciences Publications. He has spent most of his career designing software to solve specific problems. Now he’s enjoying feeding the computer data and allowing the computer to make its own decisions. What could go possibly wrong?

Presentations

AI for automation and influence in open science publishing Session

eLife’s mission is to accelerate discovery and encourage responsible behaviors in science. Daniel Ecer and Paul Shannon detail eLife’s journey in using NLP, computer vision, and similarity algorithms to find more diverse peer reviewers, apply semantics to archive content, automate the submission process, and find insights into the sentiment of scholarly content.

Thomas Endres is an IT consultant at TNG Technology Consulting in Munich. Besides his normal work for TNG’s customers, he creates prototypes with the company’s hardware hacking team, such as a see-through augmented reality device and a telepresence robotics system. In his spare time, he is working on gesture control applications, such as those for controlling quadrocopters with bare hands. He’s also involved in open source projects written in Java, C#, and all kinds of JavaScript languages. In addition to all this, he’s a lecturer at the University of Applied Sciences in Landshut. Thomas is passionate about software development and all the other aspects of technology. As an Intel Software Innovator and Black Belt, he promotes new technologies like gesture control, AR/VR, and robotics around the world. He recently received a JavaOne Rockstar award. He studied IT at the TU Munich.

Presentations

How machines learn to code: Machine learning on source code Session

Thomas Endres and Samuel Hopstock demonstrate how to apply machine learning techniques on a program's source code, covering problems you may encounter, how to get enough relevant training data, how to encode the source code as a feature vector so that it can be processed mathematically, what machine learning algorithms to use, and more.

Susan Etlinger is an industry analyst at Altimeter. Her research focuses on the impact of artificial intelligence, data, and advanced technologies on business and culture and is used in university curricula around the world. Susan’s TED Talk, “What do we do with all this big data?,” has been translated into 25 languages and has been viewed more than 1.2 million times. She is a sought-after keynote speaker and has been quoted in such media outlets as the Wall Street Journal, the BBC, and the New York Times.

Presentations

Executive Briefing: Ethical AI—How to build products that customers will love and trust Session

Susan Etlinger explores how AI fundamentally changes the relationship between people and businesses, lays out its risks and opportunities, and demonstrates emerging best practices for designing customer-centric and ethical products and services.

Guy Feigenblat is a research staff member in the AI Language Department of the Haifa Research Lab, where he leads research around machine-generated automatic document summarization. He’s also an adjunct faculty member at Haifa University. Previously, he was involved in various machine learning and AI projects that focused on developing cognitive bots that can express and predict human emotions. He has published several papers and patents. Guy holds a PhD in computer science from Bar-Ilan University, where he worked under the supervision of Ely Porat. His main academic interests include machine learning, AI, information retrieval, data mining, and data structures.

Presentations

Unsupervised automatic document summarization Session

Automatic summarization is the computational process of shortening one or more text documents in order to identify their key points. Guy Feigenblat surveys recent advances in unsupervised automated summarization technologies and discusses recent research publications and datasets. Guy concludes with an overview of a novel summarization technology developed by IBM.

Bruno Fernandez-Ruiz is cofounder and CTO at Nexar, where he and his team are using large-scale machine learning and machine vision to capture and analyze millions of sensor and camera readings in order to make our roads safer. Previously, Bruno was a senior fellow at Yahoo, where he oversaw the development and delivery of Yahoo’s personalization, ad targeting, and native advertising teams; his prior roles at Yahoo included chief architect for Yahoo’s cloud and platform and chief architect for international. Prior to joining Yahoo, Bruno founded OneSoup (acquired by Synchronica and now part of the Myriad Group) and YamiGo; was an enterprise architect for Fidelity Investments; served as manager in Accenture’s Center for Strategic Research Group, where he cofounded Meridea Financial Services and Accenture’s claim software solutions group. Bruno holds an MSc in operations research and transportation science from MIT, with a focus on intelligent transportation systems.

Presentations

Multitask networks on mobile environments Session

Bruno Fernandez-Ruiz details a unified network that jointly performs various mission-critical tasks in real time on a mobile environment, within the context of driving. Along the way, he outlines the challenges that emerge when training a single mobile network for multiple tasks, such as object detection, object attributes recognition, classification, and tracking.

Lucio Floretta is a machine learning specialist for Europe, Africa, and the Middle East at Google. He helps customers succeeding with machine learning thanks to Google Cloud.

Presentations

Cloud AutoML: Customize machine learning models with your own data (sponsored by Google) Session

Cloud AutoML is a suite of machine learning products that enables developers with limited machine learning expertise to train high-quality models specific to their business needs, leveraging Google’s state-of-the-art transfer learning and neural architecture search technology. Lucio Floretta demonstrates the power and ease of use of AutoML Vision, Translate, and Natural Language.

Giorgia Fortuna is a software developer and machine learning consultant at Machine Learning Reply, where she develops and supervises all kinds of ML projects, from building topic detectors and conversational agents to developing forecasting models. Previously, Giorgia focused on generative and unsupervised ML from a more theoretical point of view before moving to real-world cases and applications. Her background is in pure mathematics, giving her the ability to analyze and capture what is important and valuable when talking about innovation and helping her address commercial needs with machine learning technology to build intelligent systems that exploit data to increase business value.

Presentations

Unsupervised ML and fraud detection with deep neural networks Session

Many industries, including banking, financial sectors, and insurance, continuously face the problem of detecting fraudulent activities. Giorgia Fortuna explores state-of-the-art innovations in fraud detection and explains how unsupervised ML fits into the picture, focusing on signature checks and face recognition.

Christine Foster is Managing Director for Innovation at the Alan Turing Institute, where she is responsible for driving forward the institute’s goal to translate its data science and artificial intelligence research into real-world impact by forging connections between the Turing’s science activities and industry, public sector, and third sector needs to broaden the institute’s engagement with partners and extend its reach into industry. Previously, Christine advised Virgin Media on implementing machine learning models to personalize customer interactions and Liberty Global on building a world-class data science team, held leadership positions at a fintech startup, American Express, and EMI Music, built digital analytics teams, implemented predictive models, and generally worked at the intersection of data science and business. Originally from Canada, Christine started her business career as a strategy consultant with Bain & Company. She holds an MBA from INSEAD and a BA in economics from the University of Toronto.

Presentations

Beyond the contract: Effective cross-sector collaboration and the Turing-HSBC partnership Session

In 2016, the Alan Turing Institute, the UK’s new national institute for data science and AI, announced a funded strategic multiyear research partnership with HSBC. Christine Foster and Rakshit Kapoor share insights and use cases that emerged while making this ambitious and innovative cross-sector partnership work.

Natalie Fridman is the vice president of research and innovation at iSi, where she manages the development of the company’s newest and advanced technologies. Natalie’s expertise lies in algorithms for behavior analysis and prediction, decision making systems, and autonomous agents. Previously, she spent eight years at artificial intelligence and robotics research group the MAVERICK Lab; was a lecturer in the Computer Science Department at Bar Ilan University; and was an artificial intelligence team leader at Elbit Systems, where she managed all aspects of the research in artificial intelligence fields. Natalie has more than 20 publications, including highly refereed AI journal articles, conference papers, and book chapters, and has two patents for algorithms in the AI field. She holds a PhD in computer science from Bar Ilan University, where she was a president’s scholar for excellence in her studies.

Presentations

Satellite detection of moving objects in a maritime environment Session

Detection of moving vessels with satellite sensors is a challenging problem. Satellite imagery is expensive, covers a very small area, and can be acquired only at predefined acquisition opportunities. Natalie Fridman dives into this challenging problem and shares ISI's AI-based solution along with successful examples of detecting maritime vessels with ISI's satellites.

Grigori Fursin is the CTO of dividiti and a founding member of the ACM taskforce on reproducibility of the ML, AI, and systems research. Previously, he was a tenured scientist at Inria and the head of the Optimization Group at the Intel Exascale Lab. Grigori is the architect of the Collective Knowledge technology used by a growing number of universities and Fortune 50 companies to enable automatic, collaborative, and reproducible development, optimization, and codesign of efficient software and hardware for emerging ML and AI workloads in terms of speed, accuracy, energy, size, and costs. He holds a PhD in computer science from the University of Edinburgh.

Presentations

Scaling deep learning on AWS using C5 instances with MXNet, TensorFlow, and BigDL: From the edge to the cloud Session

Gaurav Kaul, Grigori Fursin, and Suneel Marthi share trade-offs and design choices that are applicable to deep learning models when training in the cloud, specifically focusing on convergence and numerical stability, which are very important for autonomous driving and medical imaging. They then demonstrate how to optimize cost, performance, and convergence using CPU spot instances in AWS.

Max Gadney runs London-based consultancy After the flood, which develops new business units and propositions for Ford Smart Mobility, Google, and Nikkei using design to navigate the 21st century deluge of data. The company excels at creating value and clarity in complex B2B domains, with design as a differentiator, allowing clients to share and validate tangible strategy.

Presentations

Executive Briefing: Why designing for trust matters Session

Max Gadney and Sabih Ali explore some of the ways designers and product teams are designing systems with transparency, trust, and privacy in mind.

Martin Goodson is the chief scientist and CEO of Evolution AI, where he specializes in large-scale natural language processing. Martin has designed data science products that are in use at companies like Dun & Bradstreet, Time Inc., John Lewis, and Condé Nast. Previously, Martin was a statistician at the University of Oxford, where he conducted research on statistical matching problems for DNA sequences. He runs the largest community of machine learning practitioners in Europe, Machine Learning London, and convenes the CBI/Royal Statistical Society roundtable, AI in Financial Services. Martin’s work has been covered by publications such as the Economist, Quartz, Business Insider, TechCrunch, and others.

Presentations

AI and financial crime Session

Martin Goodson and Mark St. John Qualter share the results of a yearlong feasibility study on the introduction of AI into the onboarding process at the Royal Bank of Scotland (RBS). Along the way, Martin and Mark share their experiences in translating this complex business process into a high-performance computational system.

Simon Greenman is cofounder and partner at Best Practice AI, a London-based AI executive advisory that helps companies accelerate their AI solutions. Simon has more than 20 years of experience leading digital transformations through technology, data science, and AI. Previously, he was a cofounder of MapQuest, one of the first internet brands; spent over 10 year as chief digital officer leading transformations of media companies for private equity; and consulted or worked for HomeAdvisor Europe, AOL, Bowers & Wilkins, and Accenture. Simon is highly active in the AI startup community. He is copresident of the Harvard Business School Alumni Angels of London, an advisor and former venture partner at DN Capital, and AI entrepreneur in residence at Seedcamp. Simon holds an MBA from the Harvard Business School and a BA in computing and artificial intelligence from the University of Sussex.

Presentations

Executive Briefing: Who is going to make money in AI? Understanding the value chain of AI Session

We're experiencing an AI gold rush. Tech giants, corporations, startups, and governments are investing billions, and headlines about AI have reached fever pitch. It's dizzying to keep track of the latest AI developments and claims. Join Simon Greenman to learn who can and who will make money in this gold rush—and who will become economic casualties along the way.

Priya Gupta is a software engineer on the TensorFlow team at Google, where she works on making it easier to run TensorFlow in a distributed environment. She is passionate about technology and education and wants machine learning to be accessible to everyone. Previously, she worked at Coursera and on the mobile ads team at Google.

Presentations

AutoGraph and distributed TensorFlow (sponsored by Google) Session

TensorFlow AutoGraph automatically converts plain Python code into its TensorFlow equivalent, using source code transformation. Brian Lee and Priya Gupta demonstrate how to distribute your training in TensorFlow easily across multiple accelerators and machines.

Sandeep Gupta is a product manager at Google, where he helps develop and drive the roadmap for TensorFlow—Google’s open source library and framework for machine learning—for supporting machine learning applications and research. His current focus is on improving TensorFlow’s usability and driving adoption in the community and enterprise. Sandeep is excited about how machine learning and AI are transforming our lives in a wide variety of ways, and he works with the Google team and external partners to help create powerful, scalable solutions for all. Previously, Sandeep was the technology leader for advanced imaging and analytics research and development at GE Global Research with specific emphasis on medical imaging and healthcare analytics.

Presentations

Building AI with TensorFlow: An overview (sponsored by Google) Session

TensorFlow is one of the world’s biggest open source projects, and it continues to grow in adoption and functionality. Sandeep Gupta and Edd Wilder-James share major recent developments, highlight some future directions, and explain how you can become more involved in the TensorFlow community.

Kristian Hammond is Narrative Science’s chief scientist and a professor of computer science and journalism at Northwestern University. His research has been primarily focused on artificial intelligence, machine-generated content, and context-driven information systems. He currently sits on a United Nations policy committee run by the United Nations Institute for Disarmament Research (UNIDIR). Kris was also named 2014 innovator of the year by the Best in Biz Awards. He holds a PhD from Yale.

Presentations

Bringing AI into the enterprise Tutorial

Even as AI technologies move into common use, many enterprise decision makers remain baffled about what the different technologies actually do and how they can be integrated into their businesses. Rather than focusing on the technologies alone, Kristian Hammond provides a practical framework for understanding your role in problem solving and decision making.

Bringing AI into the enterprise: A functional approach to the technologies of intelligence Keynote

Kristian Hammond walks you through an approach to bring AI into the enterprise, based on the functional, business aspects of AI technologies. Kristian maps out simple rules, useful metrics, and where AI should live in the org chart, laying out the route you should follow to make good on the promise of the technologies of intelligence.

Qualified in Industrial Computing and Engineering. Working 25 years with GPU’s in Application and Compute across Poweredge Server & Precision Workstation

Presentations

Efficient neural network training on Intel Xeon-based supercomputers Session

SURFSara and Intel collaborated as part of the Intel Parallel Computing Center initiative to advance the state of large-scale neural network training on Intel Xeon CPU-based servers. Ananth Sankar, Valeriu Codreanu, Damian Podareanu, and Steve Smith share insights on several best-known methods for neural network training and present results from tests performed on Stanford's CheXNet project.

Amy Heineike is the vice president of product engineering at Primer, where she leads teams to build machines that read and write text leveraging NLP, NLG, and a host of other algorithms to augment human analysts. Previously, she built out technology for visualizing large document sets as network maps at Quid. A Cambridge mathematician who previously worked in London modeling cities, Amy is fascinated by complex human systems and the algorithms and data that help us understand them.

Presentations

Natural language processing, understanding, and generation Session

When building natural language processing (NLP)-based applications, you quickly learn that no single NLP algorithm can handle the wide range of tasks required to turn text into value. Amy Heineike explains how she orchestrates natural language processing, understanding, and generation algorithms to build text-based AI applications for Fortune 500 companies.

Why we built a self-writing Wikipedia Keynote

Human-generated knowledge bases like Wikipedia have excellent precision but poor recall. Amy Heineike explains how Primer created a self-updating knowledge base that can track factual claims in unstructured text and describe what it learns in human-readable text.

Pasi Helenius is a senior business solutions manager within the Artificial Intelligence Practice at SAS.

Presentations

Business forecasting using hybrid approach: A new forecasting method using deep learning and time series Session

Business forecasting generally employs machine learning methods for longer and nonlinear use cases and econometrics approaches for linear trends. Pasi Helenius and Larry Orimoloye outline a hybrid approach that combines deep learning and econometrics. This method is particularly useful in areas such as competitive event (CE) forecasting (e.g., in sports events political events).

Chris Hillman is a London-based principal data scientist on the international advanced analytics team at Teradata. Chris is involved in the presale and startup activities of analytics projects helping customers to gain value from and understand advanced analytics and machine learning. He has over 25 years’ experience working with analytics across many industries, including retail, finance, telecom, and manufacturing. He has spoken on data science and analytics at Teradata events such as Universe and PARTNERS and industry events such as the Strata, Flink Forward, and IEEE Big Data conferences. Chris holds a PhD from the University of Dundee, where he applied big data analytics to data from the Human Proteome Project.

Presentations

The AI in fail (sponsored by Teradata) Session

Christopher Hillman explores the reasons why AI projects fail and why in some cases this is good and in others bad. Chris then explains how to avoid making the same mistakes again.

Samuel Hopstock is working toward his bachelor’s degree in computer science at the Technical University of Munich. He’s also a working student at TNG Technology Consulting in Unterföhring, where he is currently involved in the development of software in the field of machine learning with Python and Java. He is interested in any new technological developments, especially involving Android.

Presentations

How machines learn to code: Machine learning on source code Session

Thomas Endres and Samuel Hopstock demonstrate how to apply machine learning techniques on a program's source code, covering problems you may encounter, how to get enough relevant training data, how to encode the source code as a feature vector so that it can be processed mathematically, what machine learning algorithms to use, and more.

Jonny Howell is an associate on ASI Data Science’s consulting team. Previously, he was an innovation consultant at Santander UK, where he led the creation and implementation of the company’s data science strategy, a central part of the COO’s digital transformation program. Jonny has worked on a variety of different data science projects across multiple industries, from engagements focused on increasing operational efficiency to identifying new business opportunities. He holds a law degree from Bristol University. In his spare time, he loves playing the guitar. In a different life, he supported the one-hit wonder Toploader.

Presentations

AI for managers 2-Day Training

Angie Ma and Jonny Howell offer a condensed introduction to key AI and machine learning concepts and techniques, showing you what is (and isn't) possible with these exciting new tools and how they can benefit your organization.

AI for managers (Day 2) Training Day 2

Angie Ma and Jonny Howell offer a condensed introduction to key AI and machine learning concepts and techniques, showing you what is (and isn't) possible with these exciting new tools and how they can benefit your organization.

Anthony Hsu is a staff software engineer on the Hadoop development team at LinkedIn, where he works on distributed TensorFlow infrastructure. Previously, he worked on Dali, LinkedIn’s dataset access layer, and Azkaban, LinkedIn’s workflow scheduler. He has also contributed to Apache Hive and Pig.

Presentations

TonY: Native support of TensorFlow on Hadoop Session

Jonathan Hung, Keqiu Hu, and Anthony Hsu offer an overview of TensorFlow on YARN (TonY), a framework to natively run TensorFlow on Hadoop. TonY enables running TensorFlow distributed training as a new type of Hadoop application. Its native Hadoop connector, together with other features, aims to run TensorFlow jobs as reliably and flexibly as other first-class objects on Hadoop.

Keqiu Hu is a staff software engineer at LinkedIn, where he is currently working on LinkedIn’s big data platforms, primarily focusing on TensorFlow and Hadoop.

Presentations

TonY: Native support of TensorFlow on Hadoop Session

Jonathan Hung, Keqiu Hu, and Anthony Hsu offer an overview of TensorFlow on YARN (TonY), a framework to natively run TensorFlow on Hadoop. TonY enables running TensorFlow distributed training as a new type of Hadoop application. Its native Hadoop connector, together with other features, aims to run TensorFlow jobs as reliably and flexibly as other first-class objects on Hadoop.

Zhipeng Huang is an open source operation manager for Huawei. Zhipeng has been involved with various major open source communities and is now the PTL of OpenStack Cyborg project, cochair of OpenStack Public Cloud WG, and co-lead of the Kubernetes Policy WG. His current focus is open source cloud federation (OpenStack Tricircle, Ubernetes, Mesos Federation), and storage function visualization with open source software and hardware solutions.

Presentations

The last mile on democratizing AI Session

Zhipeng Huang explains how resource representation (RR) works with various intermediate representation (IR) technologies to help achieve the democratization of AI.

Lars Hulstaert is a data scientist at Microsoft. Previously, he studied machine learning at Cambridge University and Ghent University.

Presentations

Democratizing deep learning through knowledge transfer Session

Transfer learning allows data scientists to leverage insights from large labeled datasets. The general idea of transfer learning is to use knowledge learned from tasks for which a lot of labeled data is available in settings where only little labelled data is available. Lars Hulstaert explains what transfer learning is and demonstrates how it can boost your NLP or CV pipelines.

Jonathan Hung is a senior software engineer on the Hadoop development team at LinkedIn.

Presentations

TonY: Native support of TensorFlow on Hadoop Session

Jonathan Hung, Keqiu Hu, and Anthony Hsu offer an overview of TensorFlow on YARN (TonY), a framework to natively run TensorFlow on Hadoop. TonY enables running TensorFlow distributed training as a new type of Hadoop application. Its native Hadoop connector, together with other features, aims to run TensorFlow jobs as reliably and flexibly as other first-class objects on Hadoop.

Anmol Jagetia is a software engineer at Media.net, one of the biggest ad tech companies globally, which was recently acquired by a Chinese consortium for $900M USD in the third-largest ad tech deals ever. Anmol is interested in machine learning, deep and reinforcement learning, web technologies, open source software, data science, and introducing people to technology. He has spoken at a number of conferences, including the O’Reilly AI Conference in New York in 2017. He has also authored some popular open source projects such as Flatabulous, which received over 2.2K stars on GitHub and has been downloaded close to a million times. Anmol was part of HPCC as a Google Summer of Code Student and interned at the prestigious Max Planck Institute for Software Systems, Germany, and Complutense University of Madrid, Spain. He graduated from the prestigious Indian Institute of Information Technology, Allahabad. He has also published his research with the IEEE and has forthcoming papers on interesting applied aspects of machine learning.

Presentations

Building a Pokédex to recognize Pokémon in real time using TensorFlow and object recognition Session

Machine learning and object recognition have matured to the point that exciting applications are now possible. Anmol Jagetia demonstrates how to create a Pokédex that uses a camera phone to recognize the Pokémon it's looking at in real time. You'll see how to gather data, prepare your dataset, tune models, and deploy it to a mobile device, using the same tech that is used in self-driving cars.

Katharine Jarmul is a data scientist and cofounder of KIProtect, a data security and privacy company for data science workflows, based in Berlin, Germany. She researches and is passionate about ethical machine learning, natural language processing, data privacy, and information security.

Presentations

Protecting your secrets

When you train a model on private data, how much of that information does the model retain? Katharine Jarmul reviews research on attacks against models to extract training data and expose potentially sensitive information. Katharine then shares potential defenses as well as best practices when training models using private or sensitive data.

Yangqing Jia is director of engineering for Facebook’s AI platform team, which develops general-purpose open source AI solutions that serve as the backbone of Facebook AI products, such as ranking, computer vision, natural language processing, speech recognition, mobile AI, and AR. He has been influential in developing an open source deep learning software stack, many of the components of which serve as the de facto industry standard in AI. He is the creator or cocreator of Caffe, TensorFlow, Caffe2, ONNX, and PyTorch 1.0. Lately, he has been focused on the design and evolution of the AI hardware and software ecosystem and the combination of AI research and conventional wisdom of computer science.

Presentations

Applied machine learning at Facebook: An infrastructure perspective Session

Machine learning sits at the core of many essential products and services at Facebook. Yangqing Jia and Dmytro Dzhulgakov offer an overview of the hardware and software infrastructure that supports machine learning at global scale.

Rethinking software engineering in the AI era Keynote

Yangqing Jia shares a series of examples to illustrate the uniqueness of AI software and its connections to conventional computer science wisdom. Yangqing then discusses future software engineering principles for AI compute.

Marie Johnson is the managing director and chief digital officer of the Centre for Digital Business, a digital services and AI company. She explores radical innovation and questions thinking that doesn’t amplify human potential or prepare us for the future. Marie conceived and led the global codesign and cocreation effort with people with disability to deliver Nadia, the first realistic AI digital human for service delivery, which has attracted worldwide interest. An internationally experienced entrepreneur, Marie has an unparalleled track record across the public and private sector delivering significant technology, innovation and digital services transformation programs encompassing revenue, business, social services, payments, identity, immigration visa operations, and disability services. The diversity of roles covers service delivery operations, global technology strategy, chief information officer, chief technology architect, technology authority, board director and advisor, and mentor to startups. Previously, Marie led Microsoft’s worldwide public services and e-government business. The e-government and digital initiatives Marie has led have been recognized globally. These include the United Nations Public Service Award in the category “Application of ICT in government: Egovernment” for the Business Entry Point, which she led for five years. In 2006–2007, Marie was named “innovative CIO of the year for Australia” and in 2013, was named one of Australia’s “100 women of influence.” 
Marie is a board director of the Australian Information Industry Association (AIIA), an advisory member of the NSW Digital Government Advisory Panel, and a member of NZTech.

Presentations

Executive Briefing: Putting a face onto AI—After Nadia, get ready for the digital human workforce Session

What does a workforce augmented by digital humans look like? Marie Johnson shares the story of the creation of Nadia, the world’s first digital human for service delivery. Drawing on her experience developing the concept and leading the delivery, Marie presents a framework to help leaders meet exponential changes across industries augmented by digital humans, including healthcare.

Rakshit Kapoor is group chief data officer at HSBC, where he is responsible for continuing the development and implementation of HSBC’s data and analytics strategy, focusing on all aspects of the way the bank gathers, manages, analyzes, and uses the group’s data as a key competitive advantage for the bank and its customers. He is also a key player defining and executing the bank’s journey to the cloud in support its data and analytics needs. Previously, Rakshit was responsible for data strategy and analytics, business intelligence, and cognitive and robotic process automation at Travelers in the US ; held a senior role at JPMorgan Chase in New York, where he worked on the development of a world-class capability for data analytics across all the lines of business in the US consumer bank; and spent time in senior IT and data roles over a 20+ year period with TIAA-CREF, Avaya, and Oracle. Rakshit is known in the industry for his leadership and delivery of large-scale strategic IT transformation projects in the data space. He holds a bachelor’s degree in engineering from the Indian Institute of Technology, a master’s degree in information systems from George Washington University, and an MBA from the University of Maryland.

Presentations

Beyond the contract: Effective cross-sector collaboration and the Turing-HSBC partnership Session

In 2016, the Alan Turing Institute, the UK’s new national institute for data science and AI, announced a funded strategic multiyear research partnership with HSBC. Christine Foster and Rakshit Kapoor share insights and use cases that emerged while making this ambitious and innovative cross-sector partnership work.

Gaurav Kaul is an architect at AWS in the UK.

Presentations

Scaling deep learning on AWS using C5 instances with MXNet, TensorFlow, and BigDL: From the edge to the cloud Session

Gaurav Kaul, Grigori Fursin, and Suneel Marthi share trade-offs and design choices that are applicable to deep learning models when training in the cloud, specifically focusing on convergence and numerical stability, which are very important for autonomous driving and medical imaging. They then demonstrate how to optimize cost, performance, and convergence using CPU spot instances in AWS.

Driven by the passion to create a better world with AI, Daeil Kim created AI.Reverie, a simulation platform to train AI to understand the world and make it better. Daeil believes that we can create a future where issues related to food, shelter, and health can be efficiently met with the help of AI. Daeil grew up in New York City. He holds a liberal arts degree at Sarah Lawrence College, focusing on literature. An interest in medicine led him to New Mexico to research schizophrenia and to understand mental illness through artificial intelligence. He then pursued a PhD in computer science at Brown University, focusing on the development of scalable machine learning algorithms. Afterward, his interests in developing tools for investigative journalism led him to pursue a career as a data scientist at the New York Times.

Presentations

Executive Briefing: How to augment sparse training sets with synthetic data

Daeil Kim delineates the advantages of synthetic data and explains how to avoid traps that lead to dead zones and false positives. He also reviews work on simulations for synthetic data in application verticals in which it is traditionally difficult to manually acquire significant datasets.

Melinda King is a Google Authorized Trainer at ROI Training, 2017’s Google Cloud Training Partner of the Year. Melinda brings 30+ years of progressive experience with a unique combination of technical, managerial, and organizational skills. She has done solution design, development, and implementation using Google products including Compute Engine, App Engine, Kubernetes, Bigtable, Spanner, BigQuery, Pub/Sub, Dataflow, and Dataproc. Her expertise includes applying data science algorithms on big data to produce insights for optimizing business decisions. Melinda is also a Microsoft Certified Trainer with certifications for Azure, SQL Server, and Data Management and Analytics. Melinda spent 20+ years serving as a member of the US Marine Corps.

Presentations

End-to-end machine learning with TensorFlow on GCP Day (sponsored by Google Cloud) Event

Melinda King walks you through the process of building a complete machine learning pipeline, from ingest and exploration to training, evaluation, deployment, and prediction.

Jason Knight is senior technology officer at Intel, where he advances what is possible with machine learning using Intel Nervana. Jason holds a PhD in computational biology. His research included developing hierarchical Bayesian statistical models for classification of cancer tumor expression data and high-performance Markov chain Monte Carlo techniques to discover gene regulatory networks in this data using Bayesian networks. He then applied these techniques on the world’s largest database of human genomes at Human Longevity Inc.

Presentations

Deep learning at scale: A field manual Keynote

Jason Knight offers an overview of the state of the field for scaling training and inference across distributed systems from a practitioner's point of view. Along the way, Jason dives deep into available tools, resources, and venues for getting started without having to go it alone.

Cassie Kozyrkov is Google Cloud’s chief decision scientist. Cassie is passionate about helping everyone make better decisions through harnessing the beauty and power of data. She speaks at conferences and meets with leadership teams to empower decision makers to transform their industries through AI, machine learning, and analytics. At Google, Cassie has advised more than a hundred teams on statistics and machine learning, working most closely with research and machine Intelligence, Google Maps, and ads and commerce. She has also personally trained more than 15,000 Googlers (executives, engineers, scientists, and even nontechnical staff members) in machine learning, statistics, and data-driven decision making. Previously, Cassie spent a decade working as a data scientist and consultant. She is a leading expert in decision science, with undergraduate studies in statistics and economics at the University of Chicago and graduate studies in statistics, neuroscience, and psychology at Duke University and NCSU. When she’s not working, you’re most likely to find Cassie at the theatre, in an art museum, exploring the world, playing board games, or curled up with a good novel.

Presentations

The missing piece Keynote

Why do businesses fail at machine learning despite its tremendous potential and the excitement it generates? Is the answer always in data, algorithms, and infrastructure, or is there a subtler problem? Will things improve in the near future? Cassie Kozyrkov shares lessons learned at Google and explains what they mean for applied data science.

Vitaly Kuznetsov is a research scientist at Google, where he focuses on the design and implementation of machine learning tools and algorithms for time series modeling, forecasting, and anomaly detection for a variety of practical applications ranging from supply forecasting for search ads to demand estimation in networks. Vitaly has contributed to a number of different areas in machine learning, including structured prediction, ensemble learning, deep learning, and the development of the theory and algorithms for forecasting nonstationary time series. He holds a PhD in mathematics from the Courant Institute of Mathematical Sciences at New York University.

Presentations

Foundations of sequence-to-sequence modeling for time series

Vitaly Kuznetsov and Zelda Mariet compare sequence-to-sequence modeling to classical time series models and provide the first theoretical analysis of a framework that uses sequence-to-sequence models for time series forecasting.

Mounia Lalmas is a director of research at Spotify, where she leads an interdisciplinary team of research scientists working on personalization and discovery. Mounia also holds an honorary professorship at University College London. Her work focuses on studying user engagement in areas such as native advertising, digital media, social media, search, and music.

Presentations

Personalizing the user experience and playlist consumption on Spotify Session

Understanding user listening behavior is essential for personalizing music listening experiences on Spotify. Mounia Lalmas explains how Spotify uses machine learning recommenders that take into account what and how users consume playlists and the rich diversity of playlist experiences.

Danny Lange is vice president of AI and machine learning at Unity Technologies, where he leads multiple initiatives around applied artificial intelligence. Previously, Danny was head of machine learning at Uber, where he led the efforts to build a highly scalable machine learning platform to support all parts of Uber’s business, from the Uber app to self-driving cars; general manager of Amazon Machine Learning, where he provided internal teams with access to machine intelligence and launched an AWS product that offers machine learning as a cloud service to the public; principal development manager at Microsoft, where he led a product team focused on large-scale machine learning for big data; CTO of General Magic, Inc.; and founder of his own company, Vocomo Software, where he worked on General Motor’s OnStar Virtual Advisor, one of the largest deployments of an intelligent personal assistant until Siri. Danny started his career as a computer scientist at IBM Research. He is a member of ACM and IEEE Computer Society and has numerous patents to his credit. Danny holds an MS and PhD in computer science from the Technical University of Denmark.

Presentations

On the road to artificial general intelligence Session

Danny Lange discusses the role of intelligence in biological evolution and learning and demonstrates why a game engine is the perfect virtual biodome for AI’s evolution. You'll discover how the scale and speed of simulations is changing the game of AI while learning about new developments in reinforcement learning.

Ching Law is general manager of the Social and Performance Ads Department at Tencent, where he oversees engineering, products, marketing, and operations. Ching was instrumental in the development of GDT, the data and technology platform for Tencent social ads. Ching has over 12 years of experience in digital marketing and computational advertising systems. Previously, he spent eight years at Google, where he led audience and contextual targeting projects for AdSense. Ching holds a PhD in computer science from the Massachusetts Institute of Technology.

Presentations

How artificial intelligence is changing advertising in China: A conversation with Bessie Lee and Ching Law Session

Advertising in China is on the frontline of AI adoption and innovation. Join Bessie Lee and Ching Law for a conversation on how AI is changing advertising. You'll hear how China's white-hot AI advertising applications can serve as roadmaps and spark ideas in other industries and how companies like Tencent are improving performance by leveraging AI technology.

Francesca Lazzeri is an AI and machine learning scientist on the cloud developer advocacy team at Microsoft. Francesca has multiple years of experience as data scientist and data-driven business strategy expert; she is passionate about innovations in big data technologies and the applications of machine learning-based solutions to real-world problems. Her work on these issues covers a wide range of industries, including energy, oil and gas, retail, aerospace, healthcare, and professional services. Previously, she was a research fellow in business economics at Harvard Business School, where she performed statistical and econometric analysis within the Technology and Operations Management Unit and worked on multiple patent data-driven projects to investigate and measure the impact of external knowledge networks on companies’ competitiveness and innovation. Francesca is a mentor for PhD and postdoc students at the Massachusetts Institute of Technology and enjoys speaking at academic and industry conferences to share her knowledge and passion for AI, machine learning, and coding. Francesca holds a PhD in innovation management.

Presentations

A day in the life of a data scientist in an AI company Session

With the growing buzz around data science, many professionals want to learn how to become a data scientist—the role Harvard Business Review called the "sexiest job of the 21st century." Francesca Lazzeri and Jaya Mathew explain what it takes to become a data scientist and how artificial intelligence solutions have started to reinvent businesses.

Bessie Lee is founder and CEO of Withinlink, a China-based startup incubator and early-stage venture fund focused on marketing technology that supports China’s media communications industry. With more than 28 years of experience in the media communications industry in Greater China, exclusively across a number of WPP companies, Bessie draws upon an extensive network of thought leaders as well as a team of C-suite media and marketing executives, investment professionals, and entrepreneurs to offer extensive experience in business management and networks to fund and mentor startups and to create winners. Previously, she was CEO of WPP China, where she was responsible for 14,000 employees and more than $1 billion in annual revenue, and the CEO of GroupM China, where over her seven-year tenure, GroupM’s media billings in the country tripled, and GroupM was named the number one media holding group in China in terms of billings for five consecutive years from 2007 to 2012, according to RECMA. Bessie is one of the most experienced and highly respected individuals in China’s media and martech industry, and her professional success is reflected in the many hats she wears as a frequent media commentator, blogger, and a highly sought-after international public speaker at events including Cannes Lions’s International Festival of Creativity, FT Future of Marketing, and RISE. She was also the innovation jury president at the Spikes Asia Awards 2017.

Bessie has received a number of international awards for her contributions to the media industry, including the 1990 Institute’s IGNITE Award in the US (2017), China’s Business Mulan Award from China Entrepreneur (2012), Media Person of the Year from the Communication University of China (2010), Media Person with Major Contribution to the Industry by China Advertising (2010), China’s top 10 most creative media professionals by the China Economic Newspaper Association (2007), and China’s top business women leaders by the All-China Women’s Federation (2006). She was recently honored as one of the 29 most inspiring women in digital, a prestigious recognition and an important symbol for female empowerment at and by DMEXCO, The Female Quotient, and Refinery29 in 2017. She was also named a member of the World Economic Forum’s Global Agenda Council on Women’s Empowerment 2010. She is an independent director on the board of Ecovacs, the world’s second largest robotic appliance company, and cofounder of the Mobile Marketing Association in China. In 2017, she was elected cochair to lead the association on key industry issues in a market with the highest mobile internet population in the world. An avid tennis fan, Bessie is also a member of the Women’s Tennis Association Global Advisory Council, where she advises the association on advertising, publicity, and marketing in the Asia-Pacific region, particularly in China. She holds a master’s degree in communications from Illinois State University. She is based in Shanghai.

Presentations

How artificial intelligence is changing advertising in China: A conversation with Bessie Lee and Ching Law Session

Advertising in China is on the frontline of AI adoption and innovation. Join Bessie Lee and Ching Law for a conversation on how AI is changing advertising. You'll hear how China's white-hot AI advertising applications can serve as roadmaps and spark ideas in other industries and how companies like Tencent are improving performance by leveraging AI technology.

Brian Lee is a software engineer at Google Brain. He works on AutoGraph, a tool for converting plain Python code into its TensorFlow equivalent, and MiniGo, an open source replication of AlphaGoZero. Previously, Brian worked at Verily Life Sciences.

Presentations

AutoGraph and distributed TensorFlow (sponsored by Google) Session

TensorFlow AutoGraph automatically converts plain Python code into its TensorFlow equivalent, using source code transformation. Brian Lee and Priya Gupta demonstrate how to distribute your training in TensorFlow easily across multiple accelerators and machines.

Eric Liang is a second-year PhD student working in the UC Berkeley RISELab with Ion Stoica. He works on frameworks and applications for machine learning and reinforcement learning. Previously, he spent several years working on systems in industry at Databricks and Google.

Presentations

Building reinforcement learning applications with Ray Tutorial

Ion Stoica, Robert Nishihara, Richard Liaw, Eric Liang, and Philipp Moritz lead a deep dive into Ray, a new distributed execution framework for reinforcement learning applications, walking you through Ray's API and system architecture and sharing application examples, including several state-of-the art RL algorithms.

Richard Liaw is a PhD student in the BAIR Lab and RISELab at UC Berkeley working with Joseph Gonzalez, Ion Stoica, and Ken Goldberg. He has worked on a variety of different areas, ranging from robotics to reinforcement learning to distributed systems. He is currently actively working on Ray, a distributed execution engine for AI applications; RLlib, a scalable reinforcement learning library; and Tune, a distributed framework for model training.

Presentations

Building reinforcement learning applications with Ray Tutorial

Ion Stoica, Robert Nishihara, Richard Liaw, Eric Liang, and Philipp Moritz lead a deep dive into Ray, a new distributed execution framework for reinforcement learning applications, walking you through Ray's API and system architecture and sharing application examples, including several state-of-the art RL algorithms.

Simon Lidberg is a solution architect within Microsoft’s Data Insights Center of Excellence. He has worked with database and data warehousing solutions for almost 20 years in a various of industries and has more recently focused on analysis, BI, and big data. Simon is the author of Getting Started with SQL Server 2012 Cube Development.

Presentations

Executive Briefing: What’s the value of an AI center of excellence (COE)? Session

As organizations turn to data-driven strategies, there's been increasing interest in creating AI centers of excellence (COEs). Benjamin Wright-Jones and Simon Lidberg take you through the building blocks of a center of excellence and describe the value for organizations embarking on data-driven strategies.

Shaoshan Liu is the cofounder and chairman of PerceptIn, a company working on developing a next-generation robotics platform. Previously, he worked on autonomous driving and deep learning infrastructure at Baidu USA. Shaoshan holds a PhD in computer engineering from the University of California, Irvine.

Presentations

DragonFly+: An FPGA-based quad-camera visual SLAM system for autonomous vehicles Session

Shaoshan Liu explains how PerceptIn built the first FPGA-based computing system for autonomous driving.

Ben Lorica is the chief data scientist at O’Reilly Media. Ben has applied business intelligence, data mining, machine learning, and statistical analysis in a variety of settings, including direct marketing, consumer and market research, targeted advertising, text mining, and financial engineering. His background includes stints with an investment management company, internet startups, and financial services.

Presentations

Closing remarks Keynote

Program cochairs Ben Lorica and Roger Chen close the second day of keynotes.

The state of automation technologies Keynote

What technologies are ready for adoption, and how should companies and organizations evaluate automation technologies? Ben Lorica and Roger Chen highlight recent trends in data, compute, and machine learning.

Thursday opening remarks Keynote

Program cochairs Ben Lorica and Roger Chen open the second day of keynotes.

Wednesday opening remarks Keynote

Program cochairs Ben Lorica and Roger Chen open the first day of keynotes.

Angie Ma is cofounder and COO of ASI Data Science, a London-based AI tech startup that offers data science as a service, which has completed more than 120 commercial data science projects in multiple industries and sectors and is regarded as the EMEA-based leader in data science. Angie is passionate about real-world applications of machine learning that generate business value for companies and organizations and has experience delivering complex projects from prototyping to implementation. A physicist by training, Angie was previously a researcher in nanotechnology working on developing optical detection for medical diagnostics.

Presentations

AI for managers 2-Day Training

Angie Ma and Jonny Howell offer a condensed introduction to key AI and machine learning concepts and techniques, showing you what is (and isn't) possible with these exciting new tools and how they can benefit your organization.

AI for managers (Day 2) Training Day 2

Angie Ma and Jonny Howell offer a condensed introduction to key AI and machine learning concepts and techniques, showing you what is (and isn't) possible with these exciting new tools and how they can benefit your organization.

Archisman Majumdar is an assistant vice president and lead for applied AI at Mphasis Next Labs, where he conceptualizes, develops, and leads multiple products in the analytics R&D space. Archisman is responsible for the research, innovation, and go to market for the products and solutions. His areas of expertise are business analytics, machine learning, product management, and information systems research. He holds a PhD in quantitative methods and information systems from the Indian Institute of Management Bangalore (IIMB).

Presentations

Design to architecture and code using deep learning: Implications for GUI development Session

Archisman Majumdar and Jai Ganesh describe the effects of AI techniques on frontend GUI development—specifically, the use of automatically generated code and architecture from text descriptions—and share deep learning techniques for text-to-image creation and template-to-code generation, along with cloud technologies in automated deployment, management, and scaling of such applications.

Zelda Mariet is a fifth-year PhD student year in the Computer Science and Electrical Engineering Department at MIT, where she studies the theory and application of negatively dependent measures for machine learning model design and optimization. She has interned at Google (Brain, Research, and Machine Intelligence), where she studied problems related to time series prediction and determinantal point processes. Zelda is a recipient of the 2018 Google PhD Fellowship in machine learning. She holds a BS and MS from École Polytechnique in France.

Presentations

Foundations of sequence-to-sequence modeling for time series

Vitaly Kuznetsov and Zelda Mariet compare sequence-to-sequence modeling to classical time series models and provide the first theoretical analysis of a framework that uses sequence-to-sequence models for time series forecasting.

Suneel Marthi is a principal technologist at Amazon Web Services. Suneel is a member of the Apache Software Foundation and is a PMC member on Apache OpenNLP, Apache Mahout, and Apache Streams. He has given talks at the Hadoop Summit, Apache Big Data, Flink Forward, Berlin Buzzwords, and Big Data Tech Warsaw.

Presentations

Scaling deep learning on AWS using C5 instances with MXNet, TensorFlow, and BigDL: From the edge to the cloud Session

Gaurav Kaul, Grigori Fursin, and Suneel Marthi share trade-offs and design choices that are applicable to deep learning models when training in the cloud, specifically focusing on convergence and numerical stability, which are very important for autonomous driving and medical imaging. They then demonstrate how to optimize cost, performance, and convergence using CPU spot instances in AWS.

Ian Massingham is part of the technical leadership team at Amazon Web Services, where he draws on over two decades of expertise in internet technologies, technology operations leadership, architecture, and software engineering to help customers bring their ideas to life through technology. In his time at AWS, Ian has helped developers and other technical end users in companies of all sizes, from startups to large enterprises, apply cloud computing technologies, solve business problems, and exploit market opportunities.

Presentations

AI and machine learning at Amazon (sponsored by Amazon Web Services) Keynote

Ian Massingham discusses the application of ML and AI within Amazon, from retail product recommendations to the latest in natural language understanding, and explains how you can use easily accessible services from AWS to include AI features within your applications or build your own custom ML models for your own specific AI use cases.

Jaya Mathew is a senior data scientist on the artificial intelligence and research team at Microsoft, where she focuses on the deployment of AI and ML solutions to solve real business problems for customers in multiple domains. Previously, she worked on analytics and machine learning at Nokia and Hewlett-Packard. Jaya holds an undergraduate degree in mathematics and a graduate degree in statistics from the University of Texas at Austin.

Presentations

A day in the life of a data scientist in an AI company Session

With the growing buzz around data science, many professionals want to learn how to become a data scientist—the role Harvard Business Review called the "sexiest job of the 21st century." Francesca Lazzeri and Jaya Mathew explain what it takes to become a data scientist and how artificial intelligence solutions have started to reinvent businesses.

Brian McMahan is a research engineer at Joostware, a San Francisco-based company specialized in consulting and building intellectual property in natural language processing and deep learning. He is also a cofounder at R7 Speech Sciences, a company focused on understanding spoken conversations. Brian is wrapping up his PhD in computer science from Rutgers University, where his research focuses on Bayesian and deep learning models for grounding perceptual language in the visual domain. Brian has also conducted research in reinforcement learning and various aspects of dialogue systems.

Presentations

Natural language processing with deep learning 2-Day Training

Delip Rao explores natural language processing with deep learning, walking you through neural network architectures and NLP tasks and teaching you how to apply these architectures for those tasks.

Natural language processing with deep learning (Day 2) Training Day 2

Delip Rao explores natural language processing with deep learning, walking you through neural network architectures and NLP tasks and teaching you how to apply these architectures for those tasks.

Ryan Micallef is a research engineer at Cloudera Fast Forward Labs focused on studying emerging machine learning technologies and helping clients apply them. Ryan is also an attorney barred in New York and spent nearly a decade as an intellectual property litigator focused on technical cases. Ryan holds a bachelor’s degree in computer science from Georgia Tech and a JD from Brooklyn Law School. He spends his free time soldering circuits and wrenching motorcycles. He also teaches microcontroller programming at his local hackerspace, NYC Resistor.

Presentations

Federated learning Session

Imagine building a model whose training data is collected on edge devices such as cell phones or sensors. Each device collects data unlike any other, and the data cannot leave the device because of privacy concerns or unreliable network access. This challenging situation is known as federated learning. Ryan Micallef discusses the algorithmic solutions and the product opportunities.

Multitask learning in PyTorch applied to news classification Session

Multitask learning is an approach to problem solving that allows supervised algorithms to master more than one objective in parallel. Ryan Micallef shares a multitask neural net in PyTorch trained to classify news from several publications, which highlights distinct language use per publication enabled by the analysis of task-specific and agnostic representations part of multitask networks.

Marios Michailidis is a research data scientist at H2O.ai and a recent world no.1 Kaggle Grandmaster. He has worked in the marketing and credit sectors in the UK market and has successfully led multiple analytics projects for acquisition, retention, recommenders, uplift, fraud detection, portfolio optimization, and more. Previously, he was senior personalization data scientist at dunnhumby, where his main role was to improve existing algorithms and the research benefits of advanced machine learning methods and provide data insights. He created a matrix factorization library in Java along with a demo version of personalized search capability. He also held positions of importance at iQor, Capita, British Pearl, and Ey-Zein. Marios is the creator and administrator of KazAnova, a freeware GUI for quick credit scoring and data mining, which was built entirely in Java. He is also the creator of the StackNet Meta-Modeling Framework. His hobbies include competing in predictive modeling competitions. He was recently ranked first out of 465,000 data scientists on the popular data competition platform Kaggle. Marios is finishing his PhD in machine learning at the University College London (UCL) with a focus on ensemble modeling. He holds a BSc in accounting finance from the University of Macedonia in Greece and an MSc in risk management from the University of Southampton.

Presentations

H2O’s Driverless AI Session

On his journey to the top spot at Kaggle, Marios Michailidis noticed that many of the things he does to perform competitively in data challenges could be automated. Marios shares lessons learned from his Kaggle experience and shows how you can achieve competitive performance in predictive modeling tasks automatically, using H2O.ai’s Driverless AI—an AI that creates AI.

Alan Mosca is the cofounder and CTO of nPlan and a part-time doctoral researcher at Birkbeck, University of London, where his research focuses on deep learning ensembles and improvements to optimization algorithms in deep learning. Previously, Alan worked at Wadhwani Asset Management, Jane Street Capital, and several software companies as well as on several consulting projects in machine learning and deep learning.

Presentations

Harden and improve your deep learning models with targeted ensembles Session

Alan Mosca shows how any deep learning model can be improved and made more secure with the use of targeted ensemble methods and other similar techniques and demonstrates how to use these techniques in the Toupee deep learning framework to create production-ready models.

After finishing my PhD I attended the ASI Data Science Fellowship in data science and engineering and was subsequently hired by ASI with the double role of data scientist and developer advocate for the SherlockML platform.

Presentations

AI for managers 2-Day Training

Angie Ma and Jonny Howell offer a condensed introduction to key AI and machine learning concepts and techniques, showing you what is (and isn't) possible with these exciting new tools and how they can benefit your organization.

Paco Nathan is known as a “player/coach” with core expertise in data science, natural language processing, machine learning, and cloud computing. He has 35+ years of experience in the tech industry, at companies ranging from Bell Labs to early-stage startups. His recent roles include director of the Learning Group at O’Reilly Media and director of community evangelism at Databricks and Apache Spark. Paco is the cochair of JupyterCon and an advisor for Amplify Partners, Deep Learning Analytics, and Recognai. He was named one of the top 30 people in big data and analytics in 2015 by Innovation Enterprise.

Presentations

Executive Briefing: Best practices for human in the loop—The business case for active learning Session

Deep learning works well when you have large labeled datasets, but not every team has those assets. Paco Nathan offers an overview of active learning, an ML variant that incorporates human-in-the-loop computing. Active learning focuses input from human experts, leveraging intelligence already in the system, and provides systematic ways to explore and exploit uncertainty in your data.

Christopher Nguyen is president and CEO of Arimo, a Panasonic company in Silicon Valley, where he leads the development of AI platforms and solutions for the enterprise. Previously, he was engineering director of Google Apps and cofounded two other successful startups. As a professor, Christopher cofounded the Computer Engineering Program at HKUST. He holds a BS (summa cum laude) from the University of California, Berkeley, and a PhD from Stanford, where he created the first standard-encoding Vietnamese software suite, authored RFC 1456, and contributed to Unicode 1.1.

Presentations

Lessons learned implementing AI for the IoT globally at Panasonic Session

Christopher Nguyen shares lessons learned implementing multiple AI commercial projects at Panasonic. Along the way, Christopher discusses a number of use cases at various stages of implementation maturity and explains what AI really means today in enterprise products, where the key opportunities are, their impact, and key success factors in the adoption of AI across the enterprise.

Alan Nichol is cofounder and CTO of leading open source conversational AI company Rasa, where he helps create the software that enables developers to build conversational software that really works. Rasa is trusted by thousands of developers in enterprises worldwide, including UBS, ERGO, and Helvetia. Alan has years of experience building AI-powered products in industry and is the author of the DataCamp course Building chatbots in Python. He was featured in Forbes’s 30 under 30 list. Alan holds a PhD in machine learning from the University of Cambridge.

Presentations

Deprecating the state machine: Building conversational AI with the Rasa stack Session

Alan Nichol walks you through building fully machine learning-based voice and chatbots with the open source Rasa stack.

Aileen Nielsen works at an early-stage NYC startup that has something to do with time series data and neural networks. Previously, Aileen worked at corporate law firms, physics research labs, a variety of NYC tech startups, and most recently, the mobile health platform One Drop as well as on Hillary Clinton’s presidential campaign. Her interests range from defensive software engineering to UX designs for reducing cognitive load to the interplay between law and technology. She also serves as chair of the New York City Bar Association’s Science and Law Committee, which focuses on how the latest developments in science and computing should be regulated and how such developments should inform existing legal practices. Aileen is a frequent speaker at machine learning conferences on both technical and legal subjects.

Presentations

Deep prediction: A year in review for deep learning for time series Session

Deep learning for time series prediction has made rapid progress in the past few years, but performance still greatly lags that of other intelligence tasks. Aileen Nielsen offers an overview of the state of the art in 2018, covering the hottest new architectures, emerging best practices for RNN training, and long overdue standard metrics to measure and compete on neural network prediction.

Refining the Turing test in the quest for AI authenticity Session

We're in the year of the AI fake out. "Fake news" is the order of the day, as nebulous chatbots have become significant political actors. Startups peddle robotically handwritten notes and algorithmically personalized gifts for our loved ones. Soon we won't even be able to tell if a customer service agent is a real person. Aileen Nielsen asks, How should we redefine intelligence as fakes flourish?

Thomas Norrie is hardware engineer on Google’s Platforms team, where he’s helped develop multiple generations of TPUs, the deep learning hardware accelerator. He finds the intersection of hardware, software, and systems to be both fascinating and productive for solving big problems. While he’s wistful for the good ol’ days of unabated Moore’s law, he’s also excited about the opportunities to think more comprehensively about domain-specific computing.

Presentations

Acceleration with TPUs (sponsored by Google) Session

Training complex machine learning models with large amounts of data can take a very long time. Thomas Norrie explores methods for accelerating this process by distributing training across multiple accelerators and machines and leads a technical deep dive into Google Cloud’s TPU accelerators.

Gal Novik is the head of the Intel AI Lab in Haifa, Israel, where he leads a team of research scientists and engineers developing state-of-the-art machine learning algorithms and tools for researchers and data scientists. His main focus areas are deep reinforcement learning, neural network compression, and Bayesian deep learning. Previously, Gal was the founder and CTO of a fintech startup and led multiple software development teams at Microsoft delivering client and server products.

Presentations

Reinforcement Learning Coach Session

Gal Novik offers an overview of Reinforcement Learning Coach, an open source Python library that models the interaction between an agent and an environment in a modular way, making it easy for researchers to implement new reinforcement learning algorithms and for data scientists to integrate additional simulation environments modeling their business problems.

Diego Oppenheimer is the founder and CEO of Algorithmia. An entrepreneur and product developer with extensive background in all things data, Diego has designed, managed, and shipped some of Microsoft’s most used data analysis products, including Excel, Power Pivot, SQL Server, and Power BI. Diego holds a bachelor’s degree in information systems and a master’s degree in business intelligence and data analytics from Carnegie Mellon University.

Presentations

The OS for AI: How microservices and serverless enable the next generation of machine intelligence Session

Diego Oppenheimer explains why machine learning is a natural fit for serverless computing, shares a general architecture for scalable ML, discusses issues he ran into when implementing on-demand scaling over GPU clusters at Algorithmia, and provides general solutions and a vision for the future of cloud-based ML.

Larry Orimoloye is a senior business solutions manager within the Artificial Intelligence Practice at SAS.

Presentations

Business forecasting using hybrid approach: A new forecasting method using deep learning and time series Session

Business forecasting generally employs machine learning methods for longer and nonlinear use cases and econometrics approaches for linear trends. Pasi Helenius and Larry Orimoloye outline a hybrid approach that combines deep learning and econometrics. This method is particularly useful in areas such as competitive event (CE) forecasting (e.g., in sports events political events).

Andrea Pasqua is a data science manager at Uber, where he leads the time series forecasting and anomaly detection teams. Previously, Andrea was director of data science at Radius Intelligence, a company spearheading the use of machine learning in the marketing space; a financial analyst at MSCI, a leading company in the field of risk analysis; and a postdoctoral fellow in biophysics at UC Berkeley. He holds a PhD in physics from UC Berkeley.

Presentations

Forecasting at Uber: Machine learning approaches

Andrea Pasqua investigates the merits of using deep learning and other machine learning approaches in the area of forecasting and describes some of the machine learning approaches Uber uses to forecast time series of business relevance.

Amit Patankar is an engineer on the TensorFlow developer infrastructure team at Google, where he works on everything from releases and issue management to infrastructure and platform requests. Amit is passionate about the democratization of TensorFlow and is fascinated by the vast potential of applications of AI. Previously, he worked on smart home devices at Nest. Amit holds a degree in electrical engineering and computer science from UC Berkeley.

Presentations

Ready, set, go: Using TensorFlow to prototype, train, and productionalize your models (sponsored by Google) Session

Building machine learning models is a multistage process. TensorFlow's high-level APIs make this process smooth and easy, whether you are starting small or going big. Amit Patankar walks you through building, training, and debugging a model and then exporting it for serving using these APIs.

Mo Patel is an independent deep learning consultant advising individuals, startups, and enterprise clients on strategic and technical AI topics. Mo has successfully managed and executed data science projects with clients across several industries, including cable, auto manufacturing, medical device manufacturing, technology, and car insurance. Previously, he was practice director for AI and deep learning at Think Big Analytics, a Teradata Company, where he mentored and advised Think Big clients and provided guidance on ongoing deep learning projects; he was also a management consultant and a software engineer earlier in his career. A continuous learner, Mo conducts research on applications of deep learning, reinforcement learning, and graph analytics toward solving existing and novel business problems and brings a diversity of educational and hands-on expertise connecting business and technology. He holds an MBA, a master’s degree in computer science, and a bachelor’s degree in mathematics.

Presentations

PyTorch: A flexible approach for computer vision models Tutorial

Computer vision has led the artificial intelligence renaissance, and pushing it further forward is PyTorch, a flexible framework for training models. Mo Patel offers an overview of computer vision fundamentals and walks you through PyTorch code explanations for notable objection classification and object detection models.

Vanja Paunic is a data scientist in the Algorithms and Data Science Group at Microsoft London. She works on building machine learning solutions with external companies utilizing Microsoft’s AI Cloud Platform. She holds a PhD in computer science with a focus on data mining in the biomedical domain from the University of Minnesota.

Presentations

Building end-to-end computer vision solutions from pretrained deep learning models Session

Dramatic progress has been made in computer vision. Deep neural networks (DNNs) trained on millions of images can recognize thousands of different objects, and they can be customized to new use cases. Vanja Paunic and Patrick Buehler outline simple methods and tools that enable users to easily and quickly adapt Microsoft's state-of-the-art DNNs for use in their own computer vision solutions.

Recurrent neural networks for time series forecasting Tutorial

Buisnesses use forecasting to make better decisions and allocate resources more effectively. Recurrent neural networks (RNNs) have achieved a lot of success in text, speech, and video analysis but are less used for time series forecasting. Join Yijing Chen, Dmitry Pechyoni, Angus Taylor, and Vanja Paunic to learn how to apply RNNs to time series forecasting.

Dmitry Pechyoni is a senior data scientist in the Cloud AI Group at Microsoft, where he works on building end-to-end data science solutions in various domains, including retail, energy management, and predictive maintenance. Previously, he built machine learning models for display advertising Akamai and MediaMath. Dmitry holds a PhD in theoretical machine learning from the Technion – Israel Institute of Technology.

Presentations

Recurrent neural networks for time series forecasting Tutorial

Buisnesses use forecasting to make better decisions and allocate resources more effectively. Recurrent neural networks (RNNs) have achieved a lot of success in text, speech, and video analysis but are less used for time series forecasting. Join Yijing Chen, Dmitry Pechyoni, Angus Taylor, and Vanja Paunic to learn how to apply RNNs to time series forecasting.

Nick Pentreath is a principal engineer in IBM’s Cognitive Open Technology Group, where he works on machine learning. Previously, he cofounded Graphflow, a machine learning startup focused on recommendations. He has also worked at Goldman Sachs, Cognitive Match, and Mxit. He is a committer and PMC member of the Apache Spark project and author of Machine Learning with Spark. Nick is passionate about combining commercial focus with machine learning and cutting-edge technology to build intelligent systems that learn from data to add business value.

Presentations

Lessons learned building an open deep learning model exchange Session

The common perception of applying deep learning is that you take an open source or research model, train it on raw data, and deploy the result as a fully self-contained artifact. The reality is far more complex. Nick Pentreath shares lessons learned building a deep learning model exchange and discusses the future of standardized cross-framework deep learning model training and deployment.

Damian Podareanu is a co-PI for the Intel Parallel Computing Center at SURFsara and an HPC consultant for the Deep Learning for HPC and Efficient Deep Learning projects. Since 2017, he’s also been leading the Quantum Computing and Quantum Internet project. Damian focuses on optimizations and efficient scaling of machine learning algorithms. Previously, he was an AI researcher. Damian studied mathematics and computer science at the University of Bucharest, high-performance computing at the Polytechnic University of Bucharest, and artificial intelligence at the University of Groningen.

Presentations

Efficient neural network training on Intel Xeon-based supercomputers Session

SURFSara and Intel collaborated as part of the Intel Parallel Computing Center initiative to advance the state of large-scale neural network training on Intel Xeon CPU-based servers. Ananth Sankar, Valeriu Codreanu, Damian Podareanu, and Steve Smith share insights on several best-known methods for neural network training and present results from tests performed on Stanford's CheXNet project.

Ruchir Puri is CTO and Chief Architect, IBM Watson, and an IBM Fellow, who has held various technical, research and engineering leadership roles across IBM’s AI and Research businesses. Dr. Puri is a Fellow of the IEEE, and has been an ACM Distinguished Speaker, an IEEE Distinguished Lecturer, and was awarded 2014 Asian American Engineer of the Year. Ruchir has been an adjunct professor at Columbia University, NY, and a visiting scientist at Stanford University, CA. He was honored with John Von-Neumann Chair at Institute of Discrete Mathematics at Bonn University, Germany. Dr. Puri is an inventor of over 50 United States patents and has authored over 100 scientific publications on software, hardware automation methods and optimization algorithms.

Sid Reddy is chief scientist at Conversica. A recognized expert in natural language processing (NLP) and computational linguistics, Sid has designed, developed and contributed to dozens of NLP systems used in production in a wide array of use cases and industry verticals from healthcare, business intelligence, and life sciences to legal and ecommerce, including creating text-mining infrastructures from scratch at two startups and at the Mayo Clinic and founding an NLP lab at Northwestern University. Most recently, Sid was a principal applied scientist at Microsoft. His research ranges from acquiring lexical resources through distributed word vector representations learned from big data and applying them to improve state of the art in sequential labeling tasks to using functional theories of grammar for association extraction and question answering. He is a patented inventor, sought-after industry speaker, and published author with research featured in over 50 peer-reviewed publications and technical conferences. He is also an adjunct faculty member at Northwestern University and UC Berkeley. Sid holds a bachelor’s degree in computer science from the Indian Institute of Technology and a PhD from Arizona State University.

Presentations

Deep reinforcement learning: How to avoid the hype and make it work for you Session

Sid Reddy shows you how to avoid the hype and decide which use cases are the best for deep reinforcement learning. You'll explore the Markov decision process with conversational AI and learn how to set up the environment, states, agent actions, transition probabilities, reward functions, and end states. You'll also discover when to use end-to-end reinforcement learning.

Sara Robinson is a developer advocate on Google’s Cloud Platform team, focusing on machine learning. She helps developers build awesome apps through demos, online content, and events. Previously, she was a developer advocate on the Firebase team at Google. Sara holds a bachelor’s degree from Brandeis University. When she’s not programming, she can be found on a spin bike, listening to the Hamilton soundtrack, or eating froyo.

Presentations

From zero to ML on Google Cloud Platform (sponsored by Google) Session

Whether you’re new to machine learning (ML) or you’re already an expert, Google Cloud Platform (GCP) has a variety of tools to help you. Sara Robinson starts with the basics: how to use a pretrained ML model with a single API call. She then demonstrates how to customize a pretrained model with AutoML. Sara concludes by explaining how to train and serve a custom TensorFlow model on GCP.

David is a Customer Engineer on Google’s Cloud Platform team, focusing on big data & machine learning. He helps customers build awesome data driven solutions through Proof of Concepts, workshops, online content, and events. Before Google, he was associate partner and co-founder of Data Reply UK. When he’s not helping customers he can be found with his family cycling around Greenwich park, traveling or eating international food

Presentations

Machine learning at scale with Kubernetes Session

Christopher Cho details how to leverage Kubernetes and the mighty Kubernetes APIs to build a complete deep learning pipeline, from data ingestion and aggregation to preprocessing and ML training to serving. Along the way, Christopher covers Kubeflow, a Google open source solution for managing machine learning with TensorFlow in a portable, scalable manner.

Diego Saenz is a managing director in Accenture’s Digital Practice, where he leads global teams that deliver innovative digital solutions for some of the world’s biggest companies. His primary focus is on digital transformation and artificial intelligence. Diego has worked in a variety of industries including high tech, consumer products travel, automotive, and retail. Previously, Diego built and sold two successful internet businesses. He received the Inc. 500 award for leading one of America’s fastest-growing private companies.

Presentations

Executive Briefing: How AI startup and VC investment drives enterprise AI innovation Session

What do the world's most innovative and fastest growing companies have in common? They are in industries with a high level of VC funding. Accenture has analyzed five years of VC investment data to discover the AI use cases and technologies that are attracting the most money and will drive enterprise AI innovation. Diego Saenz explains where the top 10 investors in AI are placing big bets.

Ananth Sankaranarayanan is the head of AI solutions engineering in the AI Products Group at Intel Corporation, where he is responsible for coengineering AI platforms, accelerating AI performance on Intel products, and scaling solutions to worldwide customers and partners across cloud service providers, enterprise, HPC, and communication service providers. Ananth has held a number of engineering leadership roles at Intel since he first started in 2001. Previously, he led Intel’s big data analytics Solutions team. He received the Intel Achievement Award for delivering Intel’s first production high-performance computing capability and more than 30 divisional recognition awards. Ananth holds a BE in computer science and engineering and an MBA in information systems. He has been awarded two patents and has authored several technical publications.

Presentations

Efficient neural network training on Intel Xeon-based supercomputers Session

SURFSara and Intel collaborated as part of the Intel Parallel Computing Center initiative to advance the state of large-scale neural network training on Intel Xeon CPU-based servers. Ananth Sankar, Valeriu Codreanu, Damian Podareanu, and Steve Smith share insights on several best-known methods for neural network training and present results from tests performed on Stanford's CheXNet project.

Kaz Sato is a staff developer advocate on the Cloud Platform team at Google, where he leads the developer advocacy team for machine learning and data analytics products such as TensorFlow, the Vision API, and BigQuery. Kaz has been leading and supporting developer communities for Google Cloud for over seven years. He is a frequent speaker at conferences, including Google I/O 2016, Hadoop Summit 2016 San Jose, Strata + Hadoop World 2016, and Google Next 2015 NYC and Tel Aviv, and has hosted FPGA meetups since 2013.

Presentations

What is ML Ops? Solutions and best practices for applying DevOps to production ML services Session

Kaz Sato offers an overview of ML Ops (DevOps for ML), sharing solutions and best practices for bringing ML into production service. You'll learn how to combine Apache Airflow, Kubeflow, and cloud services to build a data pipeline for continuous training and validation, version control, scalable serving, and ongoing monitoring and alerting.

Robert Schroll is a data scientist in residence at the Data Incubator. Previously, he held postdocs in Amherst, Massachusetts, and Santiago, Chile, where he realized that his favorite parts of his job were teaching and analyzing data. He made the switch to data science and has been at the Data Incubator since. Robert holds a PhD in physics from the University of Chicago.

Presentations

Deep learning with TensorFlow 2-Day Training

TensorFlow is an increasingly popular tool for deep learning. Robert Schroll offers an overview of the TensorFlow graph using its Python API. You'll start with simple machine learning algorithms and move on to implementing neural networks. Along the way, Robert covers several real-world deep learning applications, including machine vision, text processing, and generative networks.

Ryan Sepassi is a senior research engineer on the Google Brain team at Google, where he works on natural language processing and reinforcement learning research and infrastructure. He’s an author and maintainer of the Tensor2Tensor library.

Presentations

Tensor2Tensor (sponsored by Google) Session

Ryan Sepassi offers an overview of Tensor2Tensor, an open source library of datasets and models and a framework for training, evaluation, and decoding, built on top of TensorFlow. Tensor2Tensor is actively used and maintained by scientists and engineers within Google Brain.

Dafna Shahaf is an assistant professor in the School of Computer Science and Engineering at the Hebrew University of Jerusalem. Dafna’s research deals with making sense of massive amounts of data. She designs algorithms that help people understand the underlying structure of complex topics, connect the dots between pieces of information, and turn data into insight. She is especially interested in unlocking the potential of the many digital traces left by human activity to understand and emulate human characteristics (e.g., creativity). Previously, she was a postdoctoral fellow at Microsoft Research and Stanford University. Dafna’s work has received multiple awards, including Best Research Paper at KDD’17 and KDD’10 and the IJCAI Early Career Award. She holds a PhD from Carnegie Mellon University.

Presentations

Accelerating innovation through analogy mining Session

The availability of large idea repositories (e.g., patents) could significantly accelerate innovation and discovery by providing people inspiration from solutions to analogous problems. Dafna Shahaf presents an algorithm that automatically discovers analogies in unstructured data and demonstrates how these analogies significantly increased people's likelihood of generating creative ideas.

Paul Shannon is head of technology at eLife, a unique collaboration between the funders and practitioners of research to improve the way it is selected, presented, and shared. He’s responsible for the technology strategy at eLife, ensuring the team is committed to openness in the products it produces to encourage broad change across the research communication landscape. Previously, he was vice president of technology at innovative digital music platform 7digital, where he grew the team and scaled the API platform to support the vastly changing music technology industry. Paul is also a regular speaker at international conferences.

Presentations

AI for automation and influence in open science publishing Session

eLife’s mission is to accelerate discovery and encourage responsible behaviors in science. Daniel Ecer and Paul Shannon detail eLife’s journey in using NLP, computer vision, and similarity algorithms to find more diverse peer reviewers, apply semantics to archive content, automate the submission process, and find insights into the sentiment of scholarly content.

Julien Simon is an artificial intelligence and machine learning evangelist for EMEA at Amazon Web Services, where he focuses on helping developers and enterprises bring their ideas to life. Previously, Julien served for 10 years as CTO and vice president of engineering at top-tier web startups, where he led large software and ops teams in charge of thousands of servers worldwide. In the process, he fought his way through a wide range of technical, business, and procurement issues, which helped him gain a deep understanding of physical infrastructure, its limitations, and how cloud computing can help. He frequently speaks at conferences and actively blogs. Julien also holds all eight AWS certifications.

Presentations

Advanced machine learning with Amazon SageMaker (sponsored by Amazon Web Services) Session

Amazon SageMaker is a fully managed platform that enables developers and data scientists to quickly and easily build, train, and deploy machine learning models at any scale. Julien Simon offers a quick overview of SageMaker. Then, using Jupyter notebooks, he dives into the more advanced features of this service.

Daniel Smilkov is a lead engineer on TensorFlow.js at Google. Previously, he worked at the intersection of visualization and machine learning, with projects like the TensorFlow Graph Visualizer and the Embedding Projector, which are part of TensorBoard, as well as new saliency techniques for neural networks. He holds a master’s degree from the MIT Media Lab.

Presentations

TensorFlow for JavaScript (sponsored by Google) Session

TensorFlow.js is the recently released JavaScript version of TensorFlow that runs in the browser and Node.js. Daniel Smilkov and Nikhil Thorat offer an overview of the TensorFlow.js ML framework and share a demo of a complete machine learning workflow, including training, client-side deployment, and transfer learning.

Kenny Song is a product manager on the Google Brain team working on TensorFlow Extended, an end-to-end production ML platform used across Google. He focuses on making ML systems easier to build, deploy, and understand. Previously, Kenny served in software engineering and product management roles at Jigsaw (an Alphabet company), Project Loon, and Google Shopping. He holds a degree in mathematics and computer science from New York University Shanghai.

Presentations

Machine learning in production with TensorFlow Extended (TFX) (sponsored by Google) Session

As machine learning evolves from experimentation to serving production workloads, so does the need to effectively manage the end-to-end training and serving workflow. Kenny Song and Quentin de Laroussilhe offer an overview of TensorFlow Extended, the end-to-end machine learning platform for TensorFlow that powers products across all of Google.

Ashok N. Srivastava is the senior vice president and chief data officer at Intuit, where he is responsible for setting the vision and direction for large-scale machine learning and AI across the enterprise to help power prosperity across the world—and in the process is hiring hundreds of people in machine learning, AI, and related areas at all levels. Ashok has extensive experience in research, development, and implementation of machine learning and optimization techniques on massive datasets and serves as an advisor in the area of big data analytics and strategic investments to companies including Trident Capital and MyBuys. Previously, Ashok was vice president of big data and artificial intelligence systems and the chief data scientist at Verizon, where his global team focused on building new revenue-generating products and services powered by big data and artificial intelligence; senior director at Blue Martini Software; and senior consultant at IBM. He is an adjunct professor in the Electrical Engineering Department at Stanford and is the editor-in-chief of the AIAA Journal of Aerospace Information Systems. Ashok is a fellow of the IEEE, the American Association for the Advancement of Science (AAAS), and the American Institute of Aeronautics and Astronautics (AIAA). He has won numerous awards, including the Distinguished Engineering Alumni Award, the NASA Exceptional Achievement Medal, the IBM Golden Circle Award, the Department of Education Merit Fellowship, and several fellowships from the University of Colorado. Ashok holds a PhD in electrical engineering from the University of Colorado at Boulder.

Presentations

AI for a better world Keynote

Industry buzz sometimes focuses on an AI future with dire unintended consequences for humanity. Ashok Srivastava draws on his cross-industry experience to paint an encouraging picture of how AI can solve big problems with people, data, and technology to benefit society.

Executive Briefing: Moving AI off your product roadmap and into your products Session

Ashok Srivastava explains how to make your organization AI ready, determine the right AI applications for your business and products, and accelerate your AI efforts with speed and scale.

Mark St. John Qualter is head of artificial intelligence for commercial and private banking at RBS. Previously, he designed and led a project that successfully transformed the Bank’s SME real estate lending proposition. Mark has held a number of other roles at RBS, including head of strategy of the Customer Solutions Group, the specialist product division of RBS’s Corporate Banking Division; director of strategy for RBS invoice finance; UK head of corporate invoice finance; managing director for North of England real estate; and regional director for corporate banking for Yorkshire and Humberside. He is currently a governor of Manchester Metropolitan University, chair of the Audit Committee, and a member of the Nominations and Governance Committee. Mark was also a member of the Yorkshire and Humberside Regional Council for the CBI. Mark holds an MBA from Manchester Business School and a BA (with honors) in Hindi and Sinhalese from the School of Oriental and African Studies at the University of London. His interests are street, documentary, and portrait photography, travel, and Krav Maga.

Presentations

AI and financial crime Session

Martin Goodson and Mark St. John Qualter share the results of a yearlong feasibility study on the introduction of AI into the onboarding process at the Royal Bank of Scotland (RBS). Along the way, Martin and Mark share their experiences in translating this complex business process into a high-performance computational system.

Rupert Steffner is the founder of WUNDER, a cognitive AI startup that is helping consumers find the products they love. Rupert has over 25 years of experience in designing and implementing highly sophisticated technical and business solutions, with a focus on customer-centric marketing. Previously, Rupert was chief platform architect of Otto Group’s new business intelligence platform BRAIN and head of BI at Groupon EMEA and APAC. He also served as business intelligence leader for several European and US companies in the ecommerce, retail, finance, and telco industries. He holds an MBA from WU Vienna and was head of the Marketing Department at the University of Applied Sciences in Salzburg.

Presentations

Sense-Infer-Act-Learn: A model for trustworthy AI Session

The increase in automated decision making, along with doubts in the quality of algorithmic decisions, has driven demand for transparency and accountability in AI. Rupert Steffner explains why the shift from black box to white box is a great opportunity to build AI models that create trust with the user and shares Sense-Infer-Act-Learn, a logical AI execution model to enable a more trustworthy AI.

Supasorn Suwajanakorn is a computer vision researcher who recently developed a technique that can synthesize a speech video of President Obama by learning from existing video footage. His earlier work includes a novel method to reconstruct a 3D face model of anyone just from their photos, which was awarded the Madrona Prize and the Innovation of the Year in 2016, as well as a software that predicts an age-progressed photo of a missing child. Previously, he was a research resident at Google Brain. Supasorn holds a PhD in computer science from the University of Washington.

Presentations

Building artificial people: Endless possibilities and the dark side Keynote

Supasorn Suwajanakorn discusses the possibilities and the dark side of building artificial people.

Angus Taylor is a data scientist in the Cloud AI Group at Microsoft, where he builds data science solutions for external customers in the retail, energy, engineering, and package distribution sectors. He holds an MSc in AI from the University of Edinburgh.

Presentations

Recurrent neural networks for time series forecasting Tutorial

Buisnesses use forecasting to make better decisions and allocate resources more effectively. Recurrent neural networks (RNNs) have achieved a lot of success in text, speech, and video analysis but are less used for time series forecasting. Join Yijing Chen, Dmitry Pechyoni, Angus Taylor, and Vanja Paunic to learn how to apply RNNs to time series forecasting.

Nikhil Thorat is a lead engineer for TensorFlow.js—a hardware accelerated machine learning library for JavaScript—at Google. His projects include the Graph Visualizer and the Embedding Projector, which are part of TensorBoard, as well as new saliency techniques for neural networks and visualizations of machine translation models. Previously, he worked on interpretability and visualization of machine learning and Google image search infrastructure.

Presentations

TensorFlow for JavaScript (sponsored by Google) Session

TensorFlow.js is the recently released JavaScript version of TensorFlow that runs in the browser and Node.js. Daniel Smilkov and Nikhil Thorat offer an overview of the TensorFlow.js ML framework and share a demo of a complete machine learning workflow, including training, client-side deployment, and transfer learning.

Jameson Toole is the CEO and cofounder of Fritz, a company building tools to help developers optimize, deploy, and manage machine learning models on mobile devices. Previously, he built analytics pipelines for Google X’s Project Wing and ran the data science team at Boston technology startup Jana Mobile. He holds undergraduate degrees in physics, economics, and applied mathematics from the University of Michigan and both an MS and PhD in engineering systems from MIT, where he worked on applications of big data and machine learning to urban and transportation planning at the Human Mobility and Networks Lab.

Presentations

Artificial intelligence at the edge Session

Machine learning and AI models now outperform humans on many tasks. However, sending sensor data up to the cloud and back is too slow for many apps and autonomous machines. Jameson Toole explains why developers seeking to provide seamless user experiences must now move their models down to devices on the edge, where they can run faster, at lower cost, and with greater privacy.

Nigel Toon is the cofounder and CEO of Graphcore. Previously, Nigel was CEO of two VC-backed silicon companies: Picochip (acquired by Mindspeed in 2012) and XMOS; cofounder of Icera (acquired by NVIDIA in 2011), a 3G cellular modem chip company, where he led sales and marketing and served on the board of directors; and vice president and general manager at Altera, where he was responsible for establishing and building the European business unit that grew to over $400M in annual revenue. He is the author of three patents. Nigel currently serves as chairman of the board of directors at XMOS and as a nonexecutive director at Imagination Technologies.

Presentations

Scaling machine intelligence with IPUs Session

Nigel Toon explains how scaling IPUs will increase the productivity of machine intelligence researchers everywhere. Join in to explore what can we do and expect from the field with vastly more compute.

Andrew Trask is a PhD student at the University of Oxford, where he researches new techniques for technical AI safety. Andrew has a passion for making complex ideas easy to learn. As such, he is the author of the book Grokking Deep Learning, an instructor in Udacity’s Deep Learning nanodegree program, and the author of popular deep learning blog i am trask. He is also the leader of the OpenMined open source community, a group of over 3,000 researchers, practitioners, and enthusiasts, which extends major deep learning frameworks with open source tools for technical AI safety.

Presentations

Building safe artificial intelligence with OpenMined

Andrew Trask details the most important new techniques in secure, privacy-preserving, and multiowner governed artificial intelligence and offers a demonstration of the OpenMined project.

Pete Warden is the technical lead of the mobile and embedded TensorFlow Group on Google’s Brain team.

Presentations

The future of ML is tiny. (sponsored by Google) Session

Pete Warden discusses the surprising effectiveness of deep learning on low-power devices.

Marc Warner is the cofounder and CEO of ASI Data Science. He founded ASI in the belief that the benefits of AI should extend to everyone and has shaped the company so that it can support organizations of all shapes and sizes to take advantage of rapid advances in this field. In the two years since founding ASI, Marc has overseen its growth to more than 50 employees and expanded its scope from a small fellowship scheme to a cutting-edge range of software, training, project, and advisory services. He has led over 50 data science projects for clients ranging from multinational companies like EasyJet and Siemens to the UK government and NHS. His work has been covered by the BBC, the Telegraph, the Independent, and many more. Previously, Marc was the Marie Curie Fellow of Physics at Harvard University, studying quantum metrology and quantum computing. His PhD research, in the field of quantum computing, was awarded the Stoneham prize and was published in Nature and covered in the New York Times.

Presentations

AI for counterterrorism Session

How can AI impact national security? Collaborating with the UK Home Office Counterterrorism Unit, ASI Data Science built a tool that removes extremist propaganda from the web. Drawing on this experience, Marc Warner discusses the role of AI in the fight against terror and explains how shared access to this technology may be part of the answer.

Fireside chat with Marc Warner and Louis Barson Keynote

Fireside chat with Marc Warner and Louis Barson

Edd Wilder-James is a strategist at Google, where he is helping build a strong and vital open source community around TensorFlow. A technology analyst, writer, and entrepreneur based in California, Edd previously helped transform businesses with data as vice president of strategy for Silicon Valley Data Science. Formerly Edd Dumbill, Edd was the founding program chair for the O’Reilly Strata conferences and chaired the Open Source Convention for six years. He was also the founding editor of the peer-reviewed journal Big Data. A startup veteran, Edd was the founder and creator of the Expectnation conference-management system and a cofounder of the Pharmalicensing.com online intellectual-property exchange. An advocate and contributor to open source software, Edd has contributed to various projects such as Debian and GNOME and created the DOAP vocabulary for describing software projects. Edd has written four books, including Learning Rails from O’Reilly.

Presentations

Building AI with TensorFlow: An overview (sponsored by Google) Session

TensorFlow is one of the world’s biggest open source projects, and it continues to grow in adoption and functionality. Sandeep Gupta and Edd Wilder-James share major recent developments, highlight some future directions, and explain how you can become more involved in the TensorFlow community.

John Wiley is a data scientist at the Wall Street Journal, where he manages a team focused on applying predictive analytics in the journal’s membership business. In collaboration with the journal’s product, design, and engineering (PDE) team, he helped develop a suite of machine learning applications enabling the dynamic targeting of paywall experiences based on a reader’s probability of subscribing. The project was recognized by the International News Media Association (INMA) as the best new paid content or subscriber initiative by a global publisher in 2018. John holds a bachelor’s degree in information systems and business analytics from Boston College’s Carroll School of Management.

Presentations

The WSJ dynamic paywall Session

Chris Boyd and John Wiley explain how the Wall Street Journal uses machine learning and a proprietary algorithm to predict the likelihood for someone subscribing, which in turn dictates the paywall experience that customer receives.

Florian Wilhelm is a data scientist at iNovex in Cologne, Germany, where he focuses on recommender systems, mathematical modeling, and bringing data science to production. Previously, he worked at Blue Yonder, the leading platform provider for predictive applications and big data in the European market, and held a postdoctoral position at the Karlsruhe Institute of Technology. Florian’s background is in mathematics. He has more than five years of project experience in the field of predictive and prescriptive analytics and big data, as well as the domains of mathematical modeli,ng, statistics, machine learning, high-performance computing and data mining. For the past few years, he has programmed mostly with the Python data science stack (NumPy, SciPy, scikit-learn, pandas, Matplotlib, Jupyter, etc.), to which he’s also contributed several extensions.

Presentations

Performance evaluation of GANs in a semisupervised OCR use case Session

Even in the age of big data, labeled data is a scarce resource in many machine learning use cases. Florian Wilhelm evaluates generative adversarial networks (GANs) when used to extract information from vehicle registrations under a varying amount of labeled data, compares the performance with supervised learning techniques, and demonstrates a significant improvement when using unlabeled data.

Benjamin Wright-Jones is a solution architect in the Microsoft WW Services CTO Office for Data and AI, where his team helps enterprise customers solve their analytical challenges. Over his career, Ben has worked on some of the largest and most complex data-centric projects around the globe.

Presentations

Executive Briefing: What’s the value of an AI center of excellence (COE)? Session

As organizations turn to data-driven strategies, there's been increasing interest in creating AI centers of excellence (COEs). Benjamin Wright-Jones and Simon Lidberg take you through the building blocks of a center of excellence and describe the value for organizations embarking on data-driven strategies.

Weiyue Wu is an investment director at the Oxford Seed Fund. Previously, Weiyue was a founding member of PerceptIn, an autonomous robotics startup company in Silicon Valley, where she brought financial, commercial, and operational experience to the team, and a senior consultant on the automotive service team at Deloitte. She holds an MBA from the Saïd Business School at the University of Oxford.

Presentations

An ecosystem analysis of the AI industry, using the case of autonomous driving Session

Does good technology equal a good product? Not necessarily. Instead of taking only technology into account, you may need to deep dive into the AI ecosystem and look at other players and factors. Weiyue Wu explains how such analysis can help in predicting AI implementation schedules, prioritizing corporate tasks, and allocating resources efficiently.

Mariya Yao is chief technology and product officer at Metamaven, a company that intelligently automates revenue growth for global companies like Paypal, LinkedIn, L’Oréal, LVMH, and WPP. She’s also editor-in-chief of TOPBOTS, the largest publication and community for business leaders applying AI to their enterprises, a Forbes writer covering the interplay of human and machine intelligence, and coauthor of Applied AI: A Handbook for Business Leaders, which she launched onstage at CES 2018.

Presentations

Executive Briefing: Organizational design for effective AI Session

Executives are being asked to "innovate with AI,” but the barriers to successful adoption for most enterprises are organizational, not technical. Mariya Yao explains why effective application of AI requires extended interdisciplinary coordination between executive and functional teams, investments in retraining your workforce, and the cultivation of an open, experimental, data-driven culture.

Alice Zimmermann looks after Google Assistant partnerships in the UK, working closely with top brands and organizations to build great content and useful local actions for the Google Assistant. She is interested to hear how you think conversational UI should be part of society and the role AI could and should play to simplify people’s lives and help them get things done. Alice has 10 years of mobile product go-to-market experience in Tel Aviv, Mexico City, and Silicon Valley. She holds degrees in business and engineering management from Stanford and neuroscience from Duke.

Presentations

The future of conversational UI Session

Fueled by the growth of messaging apps, conversational interfaces are quickly becoming an essential component of every service and product. Join Alice Zimmermann to learn how Google approaches the emerging UX challenges in its conversational agent platform. Along the way, Alice discusses the opportunities in this space and the future of conversation agents.