On SieveStack
Co-founding a biotech startup is hardly an inevitable outcome for a mathematician.
I had just defended my PhD thesis at Stanford and moved to North Carolina to start a postdoc at UNC Chapel Hill. This would probably be a good place to rant about modern academia — the years of moving from place to place, rebuilding your social life from scratch each time, until, if you’re lucky, you land a tenure-track job in some random college town, with little control over where you end up and no easy way to build anything lasting — but I decided against going into it.
Some might blame a quarter-life crisis. I blame North Carolina’s endless heat. Either way, something had shifted, and I found myself less motivated to push forward humanity’s understanding of quantum K-theory, integrable lattice models, and elliptic quantum toroidal algebras. I still love math. I do love understanding the structure of these beautiful hard sudoku puzzles, and the feeling of cracking one that many others have tried and failed. But just as much, I needed more agency and variety.
So there I was, a “math professor” (UNC Chapel Hill had invented the “teaching assistant professor” title for postdocs), teaching logic to undergrads, and spending my free time searching for a new direction to apply myself. At first my leading option was switching to AI research in the private sector, but the prospect of having a corporate job felt antithetical to my core values.
And that’s where I got lucky. You see, my dear friend Ian had been developing computational methods for drug discovery as part of his work at a UCSF lab which was his dream job. He’d been there for three years when I asked if he’d ever consider starting a company. He said he’d do it — but only if he could be the CTO, and only if he found someone else to be the CEO and do all the talking. I suddenly felt that years of lecturing and speaking at conferences made my vocal cords quite agreeable for this role.
It was the night of December 30th, 2023, at 4:10 a.m. Ian and I were walking on Market St. in San Francisco when I created a Telegram chat called “SNIK” (Slava Naprienko, Ian Knight) and sent the first messages about startup programs and name ideas. Naturally, all the names were clever derivatives of Greek words — which is, of course, how nerds go about naming things.
During the first couple of months, it became clear that it’s impossible to run a startup seriously while holding any kind of day job. That’s why I told my math department I wouldn’t be renewing my postdoc contract. I also convinced Ian to leave his dream job at the lab. But even once we started working full-time, there was an immense amount to do, and it always felt like we weren’t moving as fast as we wanted to.
That’s where I got lucky the second time. Amber, another dear friend of mine, had just sunsetted her e-commerce business and was looking for something new. Around that time, she generously offered to help us with the website. Then she built us a post-processing algorithm to analyze top-performing chemical compounds based on our early models. Then she started helping with market analysis. Then, de facto, she started working full-time. It is then Ian and I asked her to join as a co-founder.
On April 18th, 2024, the three of us incorporated the company and officially became co-founders.
The motivation behind the company probably sounds a bit too grandiose — or maybe even naive — but: we want to develop drugs for diseases that currently have no effective treatments. The driving realization is that both academic labs and traditional big pharma fail to keep up with new technology, especially one that involves a lot of GPUs. We believe that this technology can work where previous methods have failed.
We named our company SieveStack because a stack of sieves is a surprisingly accurate analogy for the drug development process, in which fewer and fewer drug candidates get promoted at each stage.
