I like solving problems at scale — and the interesting ones usually start as a mountain of messy data nobody can actually use: millions of court records, decades of government gazettes, thousands of financial filings. I build the systems that fix that — scrapers, pipelines, databases, APIs — and increasingly the AI layers on top so people can ask questions of the data in plain language.
My path has zig-zagged between public policy and startups. On the startup side, I've been a founding AI engineer at Superjoin and an AI research scientist at Jhana, building agentic products and legal & courtroom AI. On the policy side, I've released open datasets and research systems at NIPFP and XKDR Forum — and today I'm a Senior Technical Consultant at the Vidhi Centre for Legal Policy, making India's primary legal sources structured and searchable.
I hold a master's in Urban Policy and Governance from the Tata Institute of Social Sciences, Mumbai, and a bachelor's in Economics, Political Science, and Sociology from Christ University, Bangalore. My full CV is here, and my code is on GitHub.
Open-source projects turning India's public data into something you can actually use. More on GitHub.
A REST API over Indian parliamentary proceedings (questions, debates, sessions) scraped from sansad.in, plus a web frontend. The full data → API → UI stack.
Search and conversational Q&A over Kerala municipal & panchayat project records — bilingual (Malayalam + English) hybrid retrieval with page-level citations over documents that are public but unreadable at scale.
A toolkit of scrapers that build structured datasets from Indian legal and public sources — the tribunals, high courts, and the Constitution. One powered a published study on court vacations at the Bombay High Court.
Rahul Gandhi recently held a press conference on irregularities in electoral roll data provided by the Election Commission of India. Veracity of the claims aside, the methodology of data analysis is q...
Read More →