Raymond Grossman has been an avid machine and deep learning practitioner since graduating from Princeton Mathematics in 2016. He specializes in natural language processing and speech at Kensho Technologies under S&P Global, where he works as an ML Engineer. He also is an accomplished Kaggler, recently achieving overall rank 14 out of over 100,000 competitors globally under the moniker "To Train Them Is My Cause". His work on Kaggle includes winning Google's Toxic Comment Classification Challenge (1st/4551). Outside of work, Raymond enjoys playing the violin and bouldering.
Information about how different SOTA models or technologies work is readily available, but information on how to apply those models quickly and effectively is hard to come by. Covering everything from modeling tips and tricks to engineering pipelines and workflows, this talk attempts to bridge that information gap by using SK-learn to demonstrate the impact of engineering decisions on modeling pipelines.