About

The unglamorous part of AI is my favourite part.

I'm Madan Raj, a backend and AI engineer in Bangalore. I've spent four years at Mercedes-Benz Research & Development India building LLM systems that people actually rely on.

Right now I'm a Senior AI Engineer, working on agents that model knowledge graphs from technical documentation and generate test cases for enterprise systems. Before that I designed multi-agent architectures, built the retrieval pipelines behind a RAG platform that answers 10,000+ questions a month, and wired together event-driven data pipelines across dozens of microservices.

I like evaluation, observability, data quality, and the plumbing that turns a demo into something dependable. A model that scores well in a notebook is the start of the work, not the end of it. The interesting questions come after: how do you know an answer is grounded, what happens when a source document changes, and how do you find the one bad chunk behind a wrong answer?

I studied Computer Science at Nitte Meenakshi Institute of Technology, joined Mercedes-Benz as an intern in 2022, and received a Bronze Star there for GenAI work on the HR DataFactory workstream.

What I work with

Languages

  • Python
  • Go
  • SQL
  • Bash

Backend

  • FastAPI
  • Flask
  • Pydantic
  • AsyncIO
  • REST
  • Microservices
  • OAuth2

GenAI

  • LangChain
  • LangGraph
  • LlamaIndex
  • MCP
  • RAG
  • Knowledge graphs
  • Multi-agent workflows

Models

  • OpenAI
  • Azure OpenAI
  • Claude
  • Llama
  • Mistral
  • Gemini

Data

  • PySpark
  • Databricks
  • dbt
  • Airflow
  • Kafka
  • Delta Lake

Infrastructure

  • Docker
  • Kubernetes
  • GitHub Actions
  • PostgreSQL
  • Neo4j
  • Redis
  • MongoDB
  • Azure
  • AWS

MLOps

  • MLflow
  • Weights & Biases
  • DVC

Observability

  • OpenTelemetry
  • Prometheus
  • LangSmith
  • Elasticsearch
  • Kibana

Want the details? Read the full experience, browse the work, or get in touch.

esc