The first scrapers

The first step before WASP. After resigning from my job, I went on a spree of problem-statement hunting. I started off building for founders: GTM Copilot, seven agents that find leads on Reddit, GitHub, job boards and Product Hunt, and simulate which bet to run. Then, via conversations, and replacing Instagram with going to startup mixers, I got to the understanding that even lean early-stage investors are under-served.

So I turned the scrapers on the investors. One script read eleven investor subreddits for people saying they were stuck: deal flow, diligence, pro-rata, cap tables, "how do", "any tips". The answers weren't pretty: scattered Airtable sheets, broken CRMs, interns logging decks by hand. I wrote it up in September 2025. The problem isn't finding startups; it's triaging them without losing the rare outlier.

Another one scored early-stage Indian founders, many still in stealth, from public data, with a points rubric you could read line by line.