Top Decile Podcast: Anthony Atlas

Every year, roughly 175,000 students enter science competitions at the bottom level. Only about 1,750 reach the International Science and Engineering Fair and the Science Talent Search. That is one percent. Anthony Atlas was one of them. He won first place at the International Science and Engineering Fair at 17, and has a minor planet named after him to prove it. Years later, he noticed something institutional venture capital had missed entirely: the people he kept meeting from that world were founding companies, and winning.

Science Fair Fund is his pre-seed vehicle built around that observation. It backs US deep tech, AI, and biotech founders at pre-seed who were first identified at 16 through elite science competitions, a cohort that has produced 29 unicorns worth over $750 billion. On the Top Decile Podcast, Anthony sat down with Connor Sattely to explain the thesis, the data behind it, and the community he is building around it.

Watch the full episode on YouTube.

Most people picture a baking-soda volcano. Anthony sees something else. "You've got to go spend, you know, a thousand hours working on something at the cutting edge and get, you know, adults to take you seriously in your research," he said. "And you're trying to solve something that hasn't been solved."

Then comes the part that maps directly onto founding a company. "You've got to go and stand up in front of people and defend it, right. And communicate it," Anthony explained. "To go heads down and then surface and communicate and have it hold up to scrutiny, I mean these are all qualities that founders need." The winners cluster into exactly the sectors you would expect from a deep tech fund: life sciences and biotech, AI and ML from the math and CS competitors, and robotics and energy. "They start companies and the companies are amazing," he said.

Anthony did not start with a pitch deck. He started with a spreadsheet. A former founder who had gone down a PhD track in neuroscience before moving into tech, he began pulling data on science fair alumni three or four years ago. "I've got to write, you know, Python scripts and get help from friends, whatever, to pull all the data down," he said. "Because I wanted to look at it back in time and see historically if I had this fund, how would it have actually performed?"

That dataset is now his edge. "People enter into my universe when they're 15, 16, 17, whatever it is when they're competing in high school, earlier than anybody," he said. What started as 10,000 to 20,000 tracked alumni is now much larger. "We're pushing 55,000 I think now," Anthony said, "by we I mean me and all of my agents that I have working and running that are constantly pulling down data." The decision to launch came down to a single moment inside VC Lab. "There was a moment I hesitated, I almost didn't do it," he recalled. "And then I thought to myself, you know what, I'm gonna regret this if I don't give it a shot and just go all in on it."

The headline metric is stark. "The 15 percent unicorn hit rate is incredibly high," Anthony said. "That's out of the companies who raise funding." But the quality signals underneath are what he says win LPs over. On graduation, "over two thirds of them make it to and through series A," and the loss rate sits under 10 percent.

Those numbers gave him room to work. "If I can just even do this about as well as the baseline, I'm worried I would already have a world-class fund," he said. The scoring model goes further: the top quartile runs at roughly 2x those baseline figures, which is where the risk drops and the upside climbs. The LPs who lean in tend to match the thesis. "They tend to skew a little bit nerdier," Anthony said, describing quantitatively minded investors, traders, and people with deep finance backgrounds who can dig into the model themselves. Many are alumni. "This fund is sort of by alumni for alumni," he said.

The alumni network is not just a sourcing engine. "We have over 2000 alumni put their hand up, say that they're willing to help, open doors, make introductions, advise, due diligence," Anthony said. That group is already shaping deals. The first two investments are both in biotech. One uses AI to find new drug targets in an area pharma has typically overlooked, backed by a seed round from a well-known early-stage firm and led by a three-time science fair finalist out of MIT and Harvard. The second is a diagnostics platform that Anthony says can test up to 100 things in a blood sample in under seven minutes, work that would normally be sent to a lab for two weeks. Three more deals are in process. When experienced alumni ask to be introduced to a portfolio company, Anthony treats it as a buying signal. "That gives me a really good signal on these are the companies I want to back," he said.

Science Fair Fund is a pre-seed vehicle for LPs who want a defined, data-backed edge rather than a network story. Anthony's operator track record reinforces the thesis: he raised over $75 million in venture capital as a founder and helped three deep tech companies grow roughly 10x from seed to Series B, and he returns to the International Science and Engineering Fair as a Grand Awards Judge. The fund identifies exceptional technical founders earlier than almost anyone, scores them against decades of historical outcomes, and surrounds them with a 2,000-strong alumni network of operators and unicorn founders. As Anthony puts it, "if you can get to that and it's with founders that are truly remarkable, yeah, then your fund deserves to exist."

Learn more about the fund and request access on Decile Access.

Science Fair Fund is a US pre-seed fund backing deep tech, AI, and biotech founders who were first identified at 16 through elite science competitions such as the International Science and Engineering Fair and the Science Talent Search. That cohort has produced 29 unicorns worth over $750 billion. Founder and Investment Lead Anthony Atlas won first place at the International Science and Engineering Fair in 2004, has a minor planet named after him, and returns to the competition as a Grand Awards Judge. As an operator he raised over $75 million in venture capital and helped three deep tech companies grow roughly 10x from seed to Series B.