About

I’m Sourangshu Pal — an engineer and researcher working at the intersection of AI systems, machine learning, and software engineering.

This blog is where I write about building production AI systems: retrieval-augmented generation pipelines, LLM application architecture, vector search infrastructure, and the practical details that textbooks skip over.

I focus on what actually works in production — not toy examples, but real systems with real failure modes.

Topics I write about:

  • Retrieval-Augmented Generation (RAG) and semantic search
  • LLM application design and prompt engineering
  • Vector databases and ANN algorithms
  • ML system evaluation and observability
  • Python, async patterns, and API design

If you find something useful here, reach out on GitHub or LinkedIn.