01 Mercedes-Benz R&D
Knowledge graph modelling agent
A companion agent that reads technical documentation and proposes changes to a knowledge graph model.
- LangGraph
- MCP
- Neo4j
- Python
Senior AI Engineer at Mercedes-Benz R&D India · Bangalore, India
Multi-agent architectures, RAG pipelines and the event-driven plumbing underneath them. Four years of turning GenAI demos into platforms people rely on.
or press to jump anywhere
01 / Selected work
01 Mercedes-Benz R&D
A companion agent that reads technical documentation and proposes changes to a knowledge graph model.
02 Mercedes-Benz R&D
Retrieval pipelines behind an internal question-answering platform used across the company.
92% grounded answers across 10K+ queries a month
03 Personal project
An autonomous research assistant that searches the web, retrieves documents and runs code.
10K+ documents behind hybrid retrieval
04 Mercedes-Benz R&D
An agent that rewrites legacy ODI SQL as dbt models for a Databricks migration, and checks its own output.
~90% of legacy queries converted automatically
02 / Approach
RAGAS metrics, LLM-as-a-judge checks and grounded-accuracy tracking come first. Prompts get tuned against them, not by feel.
OpenTelemetry, Prometheus and LangSmith on every agent and pipeline, so a bad answer can be followed back to the chunk, prompt or service behind it.
Kafka, Airflow, health probes and CI/CD. The unglamorous parts are what turn a demo into something a team can depend on.
03 / Experience
04 / Contact
I'm happy to talk about agents, retrieval, evaluation, or getting a prototype ready for production.