Stanford MLSys Seminar Episode 66: Roman Kazinnik [Today, 1.35-2.30pm PT]

[Dan Fu]

Hi everyone,

Happy NeurIPS day! Come decompress after the deadline with your favorite MLSys Seminar. We're back with the sixty-fifth episode of the MLSys Seminar today from 1.35-2.30pm PT.

We'll be joined by Roman Kazinnik, who will talk about ML in production at Meta. The format is a 30 minute talk followed by a 30 minute podcast-style discussion, where the live audience can ask questions.

Livestream LInk: https://www.youtube.com/watch?v=eNWeFccrouI

Guest: Roman Kazinnik

Title: Machine Learning in Production: Review of Empirical Solutions

Abstract: Taking stock of ML Infra problems with potential to benefit from systematic analysis. ML currently requires running large amounts experiments to compensate for the lack of analysis. Modern AI infrastructure (major clouds) is efficient in creating, training, and deploying thousands of model. At the same time, improving production models performance, accurate estimation of models performance in production, web data relevance, risk mitigation - these are ad hoc and experiment-driven processes. Analytical analysis for Production [distributed, large-scale, rapidly changing environment] ML can help to direct and hopefully replace the empirical and manual processes.

Bio: Roman Kazinnik is working at Meta on the AI Platform team. He is an experienced computer programmer passionate about empirical and theoretical work. He created algorithms for Ads serving, deep Earth oil exploration wavefield model training, progressive image streaming, stock portfolio optimization. He is a recipient of the best paper award of the European Assoc. of Computer Graphics, and he did his Master's at Technion and Ph.D. at Tel Aviv University, Israel.

See you all there!

Best,

Dan