The Deplorable State of Deployment

Tuesday, May 9, 2017 -
11:10 am to 11:40 am

We’ve gotten pretty good at building machine learning models. From legacy platforms like SAS to modern MPP databases and Hadoop clusters, if you want to train up regression or classification models, you're in great shape. In contrast, deploying those models is a face-meltingly painful experience. This despite the fact that machine learning models are primarily only useful to a business insofar as they're deployed into operational systems that influence the business’ behavior.

In an effort to demystify machine learning, Josh will cover what he believes is the clear path forward: an approach to machine learning deployment based on an industry standard, language-agnostic, and ability to represent a broad range of algorithms. And it’s urgent — if the work of data scientists never gets deployed into operational processes, it will deliver little value and reinforce the nascent sense of disillusionment in the market concerning data science and machine learning.

VP of Product
Alpine Data
Josh Lewis is the VP of Product at Alpine Data and has ten years of experience across academia and industry in machine learning, data analysis, and user experience. Prior to joining Alpine, Josh led the frontend engineering team at Ayasdi where he built apps and APIs for the healthcare, pharmaceutical, and finance verticals, as well as Ayasdi’s domain-general data analysis and visualization software. Before joining Ayasdi, Josh was a Ph.D. student and postdoc at the UC San Diego Cognitive Science Department where he investigated the role of human perception and insight in the data analysis process. He also developed novel software for applying unsupervised machine learning algorithms called Divvy, a project that was supported by a multi-year NSF grant. Josh graduated from Pomona College with majors in Cognitive Science and Philosophy.