Selling software to hedge funds is a Catch-22: the technology is either so good that the best thing to do is start a new hedge fund, or the technology is just bad enough that funds won’t feel uncompetitive if they share it with others.
The hedge fund world is a zero-sum game. Over the past summer, I learned just what this means for Silicon Valley. Many of the financial tech tools, web scrapers, and data analytics tools out there are viewed by a hedge fund manager like this: “If this tech was any good, it would just be a hedge fund. It’s not a hedge fund, so it must be shit.”
That’s almost word-for-word what the founder of a large hedge fund told me over the phone.
So after just 4 months, my cofounders and I decided to put Apollo.ai on the shelves. Our technology helped predict a market opportunity that netted one fund $2 million. We had built a way to do automatic data integration over arbitrary data sources, using natural language processing to extract entities and create a universal ontology for a hedge fund’s data. And we had started a partnership with a large data provider to give us a unique edge.
But the gold mine for hedge funds is predictive analytics, and specifically, the construction of the predictive financial models that they build themselves. Not exactly in the purview of a tech company like ours, which sought to avoid financial modeling so we could focus on design and data analytics. I guess I realized that I’m not a finance guy.
Special thanks to the great cofounders I worked with: Chris Guthrie, Armaan Ali, and Griffin Price. And another thanks to all of our advisors and friends we met along the way, especially our Lightspeed Summer Program advisors like Jeremy Liew, and Pejman Nozad and Mar Hershenson of Pejman Mar Ventures for helping us find housing and space. I feel bad we couldn’t make it worth your time, but at least I learned more than I ever expected.
You can still check out the splash page at www.apollo.ai, and contact me if you want the pitch deck.