Data Scientist
Current+ Used discrete event simulations to forecast number of hires. This was an unusual choice and required lots of communication about how it works and how the probabilities can be used to make better decisions.+ Deployed MLflow on GCP using Docker and Kubernetes+ Used named entity recognition to reduce time taken for document reviews by more than half. This required organizing creation of training data, choosing right metric, knowledge transfers.+ Used topic modeling and classification to tag documents in a knowledge base.+ Provided consulting for operations folks around revenue forecasts, hiring requirements, probabilistic thinking.+ Rapidly prototyped interfaces to evaluate various text summarization packages.+ Did Monte Carlo simulations with Python, C, Julia, Go, Rust.