AI Engineer · Pune, India · open to relocation and remote
Ansh Kumar Singh
I build LLM systems (agents, RAG, graph retrieval) and test them on labelled question sets, so every result comes with a number and the failures behind it.
How to read this notebook
- Turn a page: swipe sideways with two fingers, press ← →, or click a folded corner.
- Jump to a chapter: pick it from Contents.
- Read a project in full: “Run the full entry” opens its case study; × or Esc brings you back to the same page.
- Turn a page: swipe sideways or tap a folded corner.
- Jump to a chapter: open Contents.
- Read a project in full: “Run the full entry”; tap × to come back.
This is a real agent loop I built. Pick a request and watch the tools fire.
Replay · GoodFoods agent
0 / 11
Logged run · gpt-oss-20bLogged runTurn the page for how I measured it →
Lab notebookAnsh Kumar SinghAI Engineer · Pune, India
LLM systems, each with its numbers and the failures behind them.
Resume (opens in a new tab)Lab notebook · endThanks for reading
If any of this is useful to your team, I’d like to hear from you.
anshkumarsingh107