10xScience, a Stanford spinout from the lab of Nobel laureate Carolyn Bertozzi, has closed a $4.8 million seed round led by Initialized Capital to tackle one of the least glamorous problems in AI-driven drug discovery: figuring out which candidates are worth pursuing in the first place.
The company is targeting a bottleneck created, ironically, by the success of generative models in biology. Systems like DeepMind's protein structure predictors now surface thousands of potential drug candidates, far more than pharmaceutical teams can meaningfully evaluate. The constraint has shifted from generation to triage.
Today the default tool for that triage work is mass spectrometry, a technique that is slow, difficult to interpret, and usable only by specialized domain experts. 10xScience is building a software layer on top of it, pairing deterministic chemistry with AI agents so that pharma researchers can work through candidate libraries at a pace that matches what generative models are producing.
Key facts
- 0110xScience. A key thread of reporting in this story.
- 02AI Biotech. A key thread of reporting in this story.
- 03Initialized Capital. A key thread of reporting in this story.
Explainability sits at the heart of the pitch. Regulators reviewing therapeutics will not sign off on a black-box verdict about what a molecule does, which means any AI layer inserted into drug discovery has to produce reasoning that chemists and agencies can audit. 10xScience is positioning its deterministic-plus-agentic approach as traceable by design.
“Everyone in the AI biotech conversation is talking about the generative side and almost nobody is building the picks and shovels layer underneath.”— Jaeden Schafer
The company is also making a bet about where value will accrue in AI biotech. Most of the venture attention has flowed to generative models that design proteins and small molecules, leaving the infrastructure for evaluating those outputs comparatively under-built. 10xScience is aiming squarely at that gap.
For Initialized Capital, the investment is a wager that picks-and-shovels software for drug discovery can turn into durable SaaS revenue inside pharma, an industry that has historically been slow to adopt outside tooling but is under increasing pressure to shorten timelines from candidate to clinic.
The broader context is that AI's impact on drug development is starting to be measured less by how many molecules models can invent and more by how quickly labs can decide which ones deserve bench time. 10xScience's seed round is a small data point, but it lines up with a shift in where the hard problems now sit.
“there's a huge bottleneck in pharma where it's not just about, you know, getting all of these candidates, but it's actually triaging them”
“regulators don't accept a black box answer on what the molecule does”
“everyone in the AI biotech conversation is basically talking about the generative side and almost nobody is building the kind of picks and shovels layer that's underneath of it”
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
