At Indegene, I designed a four-layer memory architecture for AI assistants using Python, FastAPI, Azure OpenAI, and Pinecone, helping them retain preferences and context across sessions. I also developed a recommendation engine that combines enterprise knowledge, user preferences, and conversation history.
I implemented Model Context Protocol integrations for secure connections to enterprise tools and created conversational workflows for generating structured creative briefs. In my personal projects, I built MeetingSense to turn meeting recordings into transcripts and follow-up materials, and Resume Insight AI to analyze resumes against job descriptions.
