About

Arijeet Mitra is a highly passionate and skilled Senior Software Engineer in Machine Learning at Etsy, with over 7 years of relevant experience in Distributed Systems, Distributed Machine Learning, and Big Data Systems. He holds a Master's degree in Computer Science from Carnegie Mellon University and has completed independent coursework in Machine Learning Systems from Stanford University. Arijeet's current work at Etsy involves democratizing machine learning with the Machine Learning Platform team. He is leading the initiative to reduce costs by optimizing PySpark job runtime, Kubernetes job usage, and making GCP native service cost attributions more visible in their platform, which has resulted in a 15% reduction in monthly platform costs. He has also developed a service for model profiling, which enables model developers to discover potential performance bottlenecks or cost concerns in model serving by seamlessly integrating ML profiling into their development workflows early. Additionally, he has reduced training time for vowpalwabbit models by 30% through precompute fixture data generation and p99 model latency for online inference by 10% through feature serialization and hardware optimizations. Prior to his current role, Arijeet has worked as a Research Assistant in Data Infrastructure at NASA, where he developed infrastructure for NASA's mobomo team for metered power and data analytics for energy optimization. He has also worked as a Graduate Student Researcher in Security Infrastructure at NSA, where he built a pipeline for performing file forensics in Cloud by leveraging AWS Compute. His work involved developing a data pipeline with components for md5/sha256 hash checking, fuzzy hash using ssdeep, and pattern anomaly detection using Yara rules. He also implemented a MySQL cluster as an evolving cache for improving performance, which was the central piece of the entire architecture. Arijeet's technical expertise includes ML, Software Engineering, infra, GCP, Big Data, Kubernetes, Docker, Microservices, Salesforce, Integration, and Research Scientist. He is highly passionate about data and interested in building systems around it to solve critical problems as he grows as an engineer. His heart is in his work, and his work is in progress.