At Google Research/AI, I worked on building foundational models, including LLMs and Transformers, advancing state-of-the-art neural models in NLP with a focus on practical applications like long document understanding, question answering, retrieval, ranking, and attribute extraction. Our research led to several publications in top conferences, and some of my open-source models have seen broad adoption, with over 1 million downloads on Hugging Face. Google Scholar: https://scholar.google.com/citations?user=BEh7J1MAAAAJ&hl=enPreviously, as a founding engineer at ThoughtSpot, helping it grow from inception to unicorn status with hands-on involvement in several 0-1 initiatives, including building the middleware stack and creating the industry’s first automated AI insights dashboard for large-scale analytics systems.I bring a blend of practical and research expertise across systems design, deep learning, NLP, foundational models (LLMs / Transformers, multi-agent AI systems), ranking, and retrieval—and enjoyed winning several hackathons during grad school.