At Morgan Stanley, I built LE Insights, a full-stack GenAI platform for searching legal entities and their precedents. It cut legal entity closure information collection from about a week to 10–30 minutes.
I designed its RAG architecture on Snowflake Cortex AI, including document chunking, embeddings, a vector database, semantic search, and ingestion of JSON and binary sources. I also developed the entity search, entity view, and integrated AI chat UI.
I owned Regal AutoSys Floweaver from brainstorming and design through production, building it with FastAPI and Angular. The platform integrated Morgan Stanley’s in-house GenAI products, visualized workflow dependencies, and improved team productivity by 70% while cutting process time by 40%.
