Project 01, AI learning platform
LuminaIQ
Make every page answer back.
A study companion that turns dense PDFs into conversations, quizzes and notes.
Shyamnath Sankar
Finding the signal
000
I build AI products end to end: retrieval systems, agents and the interfaces people use every day.
ProjectsWhat I build: RAG pipelines, Agent graphs, MCP servers, Browser agents, Vector search, Flutter apps, Next.js, FastAPI, Product strategy.
Most AI demos stop at the demo. I keep going until someone depends on it: a student revising the night before an exam, a college reading its own feedback, a developer handing a browser to an agent.
I’m Shyamnath, a Computer Science & Business Systems student and co-founder of Genrecai, where I lead product and AI engineering. Most of my work is turning messy inputs (PDFs, spreadsheets, open web pages, raw databases) into answers people can trust.
More about meFive products, 2025 to now
Project 01, AI learning platform
Make every page answer back.
A study companion that turns dense PDFs into conversations, quizzes and notes.
Project 02, Open-source browser agent
Hand your real browser to an agent.
A Chrome side-panel agent that any model can drive, plus an MCP bridge for coding agents.

Project 03, Agentic database analysis
Ask the data. See the answer.
Connect a database, ask in plain English, get SQL, charts and models back.

Project 04, Cross-platform AI app
Every model. One room.
One focused workspace for GPT, Claude, Gemini, Llama and Groq, with your own keys.
Project 05, Campus feedback system
Listen. Then act.
A campus-wide feedback system that turns student voice into decisions.
Smaller tools and experiments, all open on GitHub. Some became products; one was retired so SelectStar could exist.