- This event has passed.
Proof of Concept to Model as a Service (MAAS): Learning to Productionize On-Premise
September 18, 2018 @ 6:00 pm - 8:00 pm
While Wells Fargo had known about the H2O platform for a while, we were limited to Proof of Concept/Technology (POC/T) projects. We knew eventually we would have to take a POJO or MOJO (H2O Optimized Model) and bring it into production. One aspect of working in machine learning at Wells Fargo is realizing that cloud and even packaged systems and libraries can be difficult to use for a variety of security regulation challenges. This meant a lot more design and planning was needed than we had envisioned. In this talk, I will take you through the processes we had to go through to design the workflow, data pipelines, historical feature generation, and MAAS related systems, all on premise. I’ll also discuss the processes we went through to bring models to production, and other technical considerations to bring a MOJO into a MAAS architecture that is able to be used internally.
If you cannot attend live, this event will also be Livestreamed.
http://livestream.com/accounts/23925505/events/8340609
About Metis
Metis (thisismetis.com) accelerates careers in data science by providing full-time immersive bootcamps, evening part-time professional development courses, online resources, and corporate programs based in Seattle, New York, Chicago, and San Francisco.
Learn more about Metis at http://thisismetis.com (http://thisismetis.com/)
MSpeaker’s Bio:
Matthew Schlachtman has worked at Wells Fargo for five years and has eight years of data science experience overall. Matthew’s role at Wells Fargo has had many different terms, including data scientist, data engineer, and machine learning engineer, as well as the less defined UAT engineer or data wrangler. Lately, his background in systems architecture, and experience in Machine Learning and H2O have led him to be a lead project architect in the design of a MaaS Architecture at Wells Fargo. Matthew studied autonomous robotics in grad school and eventually transitioned to using his Computer Science and Artificial Intelligence background for non-physical applications.
