Summary
Vipul Raheja is a research scientist specializing in NLP, large language models, and conversational agents, with 14 years of experience bridging academia and industry. Based in San Francisco, he has driven impactful AI work at Grammarly, x.ai, IBM Research, and Google DeepMind, focusing on LLM personalization, RLHF/RLAIF, retrieval-augmented generation, and tool-augmented agents. His projects span multilingual text revision, AI-written content detection, and document-level understanding, including end-to-end NLP pipelines for information extraction and entity resolution in scheduling contexts. He combines a solid academic foundation from IIIT Hyderabad and Columbia University with a hands-on open-source track dating back to Google Summer of Code, including OSSIM bindings. Vipul’s career reflects a rare ability to turn cutting-edge NLP research into scalable, user-centric products that work across languages and domains.
14 years of coding experience
14 years of employment as a software developer
Columbia University in the City of New York