• Agentic AI developer: I design multi-agent systems with LangGraph, LangChain and Claude, covering tool use, human-in-the-loop flows, planning and stateful workflows.
• I connect agents to real systems with MCP (Model Context Protocol): infrastructure, monitoring and operational tooling.
• I build RAG and semantic search at scale, including an embeddings + pgvector search over millions of podcast episodes.
• I build voice AI: local Whisper speech-to-text and an autonomous interview bot that speaks in live Google Meet calls.
• I have a strong Cloud & DevOps background (AWS, Docker, Kubernetes, Terraform, CI/CD), so the agents I build actually ship.
• 2+ years at Feedspot shipping production integrations with Claude, OpenAI, Amazon Bedrock and Amazon Q, plus AI-assisted data pipelines that run about 30% more efficiently.

Click any badge to verify the credential.

Multi-agent investment research on live Indian market data, with cited reports and portfolio rebalancing.	Live meeting notes and an AI interviewer that asks resume-based questions, scores answers and speaks in Google Meet.
Agentic customer support assistant with RAG, tool calling and human-in-the-loop refunds.	Free AIF-C01 notes, cheat sheets and Bedrock labs, learning in public.

AI: Hey cat, what are you waiting for? I've already calculated every move.
Cat: "A good player makes the move when it's time for checkmate."