[ AI Engineer — Pune, India ]
I build knowledge graphs, agentic workflows and production RAG — the parts of AI that have to be right, not just fluent.
[ 01—About ]
I turn scattered knowledge into systems you can ask questions of — and trust the answers you get back.
For four years at Dassault Systèmes I have worked on both halves of that problem. First the interface: fast, well-tested web components for an enterprise platform used by more than 2,000 people. Now the intelligence: LLM assistants, tool-calling agents that pause for a human before they act, and MCP integrations that let those agents work inside real engineering software.
Outside work I am finishing an M.Tech in AI & ML at BITS Pilani, where I built GraphMind — hand it documents, code, audio or an entire repository and it gives back a knowledge graph you can explore, query, and pass to other agents. The parts I enjoy most are the unglamorous ones: evaluation, latency, and being able to explain why a model said what it said.
[ 02—Work ]
Systems I have designed end to end. Each one has a write-up covering architecture, trade-offs and measured results.
Turn anything into a knowledge graph you can ask questions of.
A multimodal knowledge-graph platform: documents, code, audio, video and whole repositories become queryable Neo4j graphs, with GraphRAG chat and an MCP server so other agents can use them.
An agent that crawls Substack each week and hides one great essay for every visitor.
A local-LLM agent that walks Substack's recommendation graph, judges thousands of essays on philosophy, psychology, critical thinking, money, science and deep space, verifies every quote word-for-word, and publishes a rotation of up to 500 that this site hands out one catch at a time.
[ 03—Experience ]
Apr 2025 — Now
Dassault Systèmes
Jul 2022 — Mar 2025
Dassault Systèmes
2024 — Sep 2026
BITS Pilani
2022
Savitribai Phule Pune University
Recognition
[ 04—Contact ]
Open to AI engineering roles, and always up for a conversation about retrieval, agents or knowledge graphs.