Improve Your Data Science ROI with a Portfolio and Risk Management Lens
Who is this presentation for?CxO, Data Science & Business Leadership, Data Scientists
Prerequisite knowledge1. Basic familiarity with data science processes and concepts 2. Helpful but not required, some frustration managing or leading data science programs
What you'll learn
Building a data science capability requires tremendous investment in people, new processes and technologies. While the return on such investments can be large, such returns are not immediately observable nor even measurable. Such issues often leave business executives in the position of making data science investments more on faith than on evidence. As the data science industry matures, both data science leaders and their executive stakeholders need to develop robust and better frameworks to deliver business impacts and justify spending levels. This talk identifies multiple levels of systemic risk that data science projects face and introduces a portfolio-based approach to hedge against such risks and thus increase the likelihood of creating value through data science.
Most business projects face both Hypothesis and Execution risks, where the former is defined by the business’s ability to identify the right opportunities, and the latter by the ability to execute well to capitalize on a such opportunities. These risks extend to data science projects, but within this domain is another, unique risk, which we call Signal risk. Signal risk is the underlying uncertainty in being able to acquire the right data related to a problem and then finding exploitable patterns within it (analogous to the concept of statistical power, but extended to broader business concerns). All together, standard business and product development processes fail to incorporate the compounded effects of these three risks. We address the issue in this talk and present concrete strategies to better hedge against such risks to improve the success rate of data science programs.
While this talk is aimed at business stakeholders, we advocate that data scientists and data science leadership must be equally vested in proving the value of their work, and thus should be included in the target audience set. We present techniques and strategies that help mitigate the aforementioned risks, learned through years of consulting and data science leadership. Ultimately, the right approach requires negotiation and engagement from all stakeholders, and our talk provides a starting basis to have such a conversation.
Brian Dalessandro is the head of data science at SparkBeyond, a research and consulting platform that accelerates discoveries and insights. Brian is also an active professor for NYU’s Center for Data Science graduate degree program. Prior to SparkBeyond, Brian has built and led data science programs for several NYC tech startups, including Zocdoc and Dstillery. A veteran data scientist and leader with over 15 years of experience developing machine learning-driven practices and products, Brian holds several patents and has published dozens of peer-reviewed articles on the subjects of causal inference, large-scale machine learning, and data science ethics. When not doing Data Science, Brian likes to cook, create adventures with his family and surf in the frigid north Atlantic waters.
Leave a Comment or Question
Help us make this conference the best it can be for you. Have questions you'd like this speaker to address? Suggestions for issues that deserve extra attention? Feedback that you'd like to share with the speaker and other attendees?
Join the conversation here (requires login)
For conference registration information and customer service
For more information on community discounts and trade opportunities with O’Reilly conferences
For information on exhibiting or sponsoring a conference
View a complete list of Strata Data Conference contacts