At Pravidhi Technical Wing, I shipped and owned tools solo, end to end, for 1,000+ active campus users. I cut query runtime by 50% on real usage by exposing slow paths.
At Google Students India, I taught 900+ student developers to build LLM agents and RAG pipelines from scratch, favoring fast working builds over perfect theory.
I built ET Investor Intelligence, a multi-agent financial platform that retrieves market data, reasons across stocks, trades, news, and sentiment, and ranks stocks. The project placed in the top 20 nationally at the Economic Times AI Hackathon.
I also built TrialGo, a clinical trial matching platform with a multi-agent RAG workflow for trial discovery and patient–trial eligibility ranking; it was runner-up at the Cognizant Technoverse 2026 National Hackathon. TruthScope, my fake news detection agent, was also a runner-up and combines a Chrome extension with a Flask backend to check evidence, analyze bias, and flag AI-generated media.
