Summary
Aritra Ghosh is a Staff Research Scientist at Meta with a decade of experience building production-grade ML systems, specializing in LLM systems, retrieval-augmented generation, long-context inference, and robust learning. He combines deep research rootsโa PhD from UMass Amherst focused on differentiable data selection, sequential/adaptive modeling, and robustnessโwith hands-on engineering experience from Microsoft and Adobe on real-time bidding and ads relevance. At Meta he designs scalable RAG and decoding systems for LLM agents, emphasizing latency, long-context efficiency, and end-to-end performance under real-world constraints. He has led probabilistic and decision-support modeling for advertiser guidance, bridging statistical ML research with deployable infrastructure. Based in Bellevue, WA, he brings both academic rigor and production discipline to challenging inference and retrieval problems, often optimizing for constrained compute and noisy supervision scenarios.
10 years of coding experience
7 years of employment as a software developer
Artificial Intelligence, Artificial Intelligence at Indian Institute of Science (IISc)
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at University of Massachusetts Amherst