Simon Kohl turned an internship at Google DeepMind into a key role in building AlphaFold2, the Nobel Prize-winning AI model that predicts the structure of proteins.
Now, Kohl is venturing out on his own. His startup, Latent Labs, launched Wednesday with $50 million following about 18 months in stealth. The biotech is building AI models to design and optimize proteins. Despite being a 34-year-old first-time CEO, Kohl voiced confidence that his four years at DeepMind, the Google-owned AI research powerhouse, have prepared him for the job.
“I’ve learned how to cook the meal in AI for biology, and I know how to do this,” Kohl said. “We are now working on models that allow us to go way beyond, we can generate from scratch proteins that don’t exist in nature. That’s what we think is the next big thing that will unlock a lot of impact.”
The London- and San Francisco-based startup is nearing 15 employees, Kohl said, with plans to sell its AI capabilities to biopharma customers instead of building its own drug pipeline. Latent’s initial focus is proteins, but Kohl hinted that more of the biological world could be in the startup’s plans.
“We are starting in the area of protein design,” Kohl said. “The vision is grander. I think we can expand from there, and we will find over time many other areas beyond the molecular interactions level will be steerable with generative models.”
But Kohl provided few details behind that vision. Latent has not publicized any details on its AI research in journal articles, preprints or whitepapers. In a recent interview with
Endpoints News
, Kohl declined to provide specifics on Latent’s AI models, their performance, revenue or number of biopharma customers.
It’s unclear when — or if — Latent will share more on what it’s working on. Kohl also declined to say if Latent plans on publishing or presenting any of its research in 2025.
“We may choose to reveal more, but we haven’t committed to that,” he said.
Latent’s launch shows how AI-focused startups can buck tradition in biotech. Most biotechs are formed around a molecule, research paper or key intellectual property. Instead, Latent’s investors are betting on AI talent in Kohl and Alex Bridgland, another ex-DeepMind developer of AlphaFold, to figure it out.
Kohl’s work at DeepMind may hint at what Latent could be working on. After AlphaFold2, he worked on protein design. DeepMind
posted a preprint in September on AlphaProteo
, a
de novo
design model for protein binders that Kohl said he helped steer while there.
Kohl left DeepMind and started Latent Labs in June 2023, while DeepMind was building out its biotech spinout,
Isomorphic Labs
, which is focused on small molecules instead of protein-based drugs.
“I felt quite strongly that this deserved its own focus and a nimble company working on generative protein design alone would be able to succeed and probably have more impact,” Kohl said.
Latent is entering a crowded arena in AI models for protein design. New models are coming out on a regular basis, with other notable startups in the space including
Xaira Therapeutics
,
Generate:Biomedicines
,
Profluent
,
Cradle Bio
, and
Nabla Bio
.
Kohl compared competition in the field to launching rockets. Different types of rockets can all reach orbit, he said, adding that Latent has seen “great results” internally with its own models. (He declined to provide specifics about those great results.)
Latent raised a seed round of nearly $10 million shortly after it was incorporated, Kohl said, with the first money from Pillar VC, Kindred Capital and 8VC. The Series A round closed in December, which was co-led by Radical Ventures and Sofinnova Partners.