- 🤖 I spend most of my time thinking about agents — how they fail, how to evaluate them, and how to ship them to people who don't care how they work (they just want it to work)
- 🧠 I've been doing AI PM for a couple of years now. Before that I thought I wanted to be a data scientist. I was wrong, and I'm glad
- 🔬 A chunk of my work has been in evals — building frameworks to test, debug, and measure AI agents before they break in production (ideally before, anyway)
- 🛒 The rest lives in e-commerce — marketplaces, catalogs, inventory, the unglamorous infrastructure that makes products actually show up when you search for them
- ✍️ I write at AGI in Progress — working through ideas on AI, products, and what it means to build in this moment
- 📺 Sometimes I make videos. They live at MoodyMind and are deeply unrelated to AI
- 🎓 CS from IIIT Vadodara — which means I have strong opinions about data structures and even stronger opinions about tabs vs spaces (tabs, obviously)
for building agents
agent primitives & protocols
the LLMs I talk to most
for RAG (the boring but important stuff)
for knowing when things are broken
for shipping it
how I actually write code
|
|
|
no meme here. just go touch grass. or read my substack. same thing.