Presented By
O’Reilly + Cloudera
Make Data Work
29 April–2 May 2019
London, UK

Building custom machine learning models for production, without ML expertise

Alicia Williams (Google)
11:3012:00 Tuesday, 30 April 2019
Business Analytics and Visualization
Location: Capital Suite 12
Average rating: ***..
(3.75, 4 ratings)

Machine learning is a powerful tool for solving business problems such as vision, speech recognition, language translation, and more. But in order to leverage machine learning, many businesses invest heavily to bring in researchers and data scientists that build and train models to solve their most critical business challenges.

However, you don’t need to be an expert to bring ML to your business. Instead of building models from scratch, companies can start with pretrained machine learning models to tackle common tasks. Then, they can customize these models in the case of domain-specific needs using automated machine learning (AutoML)—all without requiring deep ML expertise.

Alicia Williams explains how two media companies followed this path to organize content and make it accessible around the world. Along the way, she details the business problems they solved with ML, demonstrates the ease of use of the tools themselves, and shows the value that ML has brought in each case.

Photo of Alicia Williams

Alicia Williams

Google

Alicia Williams is an advocate for Google Cloud. Previously, she spent six years as a program manager; through building, managing, and measuring programs and processes, she fell in love with data science. Known to hang out in spreadsheets surrounded by formulas, she also uses machine learning, SQL, and visualizations to help solve problems and tell stories.