Coopsight

ML-powered startup matching platform. 500 startups signed up, Sberbank partnership in discussions. Built and led a team of 9 (2019-2021).

500 startups signed up. We were in talks with Sberbank, Russia’s largest bank, to apply our matching system to their portfolio. And then COVID froze the entire market we were building for.

But let me back up.

It’s 2019. I’m a sophomore at NYU Shanghai, a Sino-US research university designed to produce global citizens, and my best friend George just came back from working as a junior analyst at VCs overseas. We’re living in China, watching WeChat and Alipay do something that doesn’t exist in the West: ecosystem-level integration where services compound on each other. Payments feed into social, social feeds into commerce, commerce feeds into logistics. Everything is connected.

We started looking at the startup world and saw the opposite. Startups in the same accelerator that never talk to each other. Portfolio companies with complementary tech that never get introduced. VCs pouring billions (SoftBank’s Vision Fund era) without any ecosystem strategy connecting their bets together.

70% of startups fail. We thought a meaningful chunk of them fail because they’re isolated. Missing the distribution partner, the tech integration, the customer base overlap that could change their trajectory. What if you could surface those connections with data?

The best partnerships aren’t found. They’re computed. Humans miss patterns at scale. Systems don’t.

The three of us (George as CEO, Erol as CTO, me as COO) built it. We scraped Crunchbase, PitchBook, and LinkedIn for company data. Startups could also upload their own pitch decks, which we OCR’d and ran through keyword extraction. We built a global keyword bank modeled on the partnership patterns we’d seen in China and used it to score complementary technology relationships.

The product was basically Tinder for startups. Matches ranked highest to lowest by compatibility score. Both sides swipe, mutual match opens a conversation. I handled the full product surface alongside managing the team, financials, and operations. I also set the design groundwork for the brand identity across the platform. We successfully presented to prominent VCs and accelerators, showing how portfolio companies could be matched to create synergies.

Then COVID hit and the market we were serving effectively froze. We made the call to close the chapter and graduate.

What stayed with me: the data matching work. Scraping, OCR, keyword mapping, scoring compatibility across dimensions. That was my first real exposure to building systems that surface insights humans miss at scale. More importantly, I learned how to lead a cross-functional team of 9 across design, finance, product, and marketing while simultaneously building the product itself. The same thinking now drives the data infrastructure I build at Blue Origin and the intent mapping at Mosslayer.