AI is spitting out more potential drugs than ever before. This startup wants to find out which ones are important.

The biggest impact of AI in science is Google DeepMind’s use of a deep learning model to predict the complex structures of proteins – molecules that drive almost every process in living cells.

But as AI models continue to spit out more people for potential treatments, a bottleneck is emerging: actually defining all those candidates, for testing and mass production.

That’s the goal of 10x Science, a startup founded in December 2025 that announced a $4.8 million seed round today, led by Initialized Capital and supported by Y Combinator, Civilization Ventures, and Founder Factor. Its three founders are David Roberts and Andrew Reiter, experienced biochemists, and Vishnu Tejas, a serial inventor with expertise in computer science and AI models.

“When a biopharma is trying to develop a drug candidate, they have all these great predictive tools,” Roberts told TechCrunch. “You can add as many candidates as you want at the top of the funnel, but they all have to go through this screening process. Everything needs to be measured.”

Understanding protein structure is key to researchers developing biologic drugs, which are produced in living cells and use complex design to target diseases and conditions specifically. For example, they can be designed to target specific cells, such as Keytruda, a popular drug sold by Merck that helps the immune system identify and attack cancer.

The three inventors of 10x worked together in the Stanford lab of Nobel laureate Dr. Carolyn Bertozzi, where they studied the communication between cancer cells and the immune system, and were frustrated by their inability to understand exactly what was happening at the molecular level.

The most accurate way to examine molecules is to use a sophisticated technique called mass spectrometry, a method of determining their atomic composition by measuring them in an electric field. The relatively new method produces complex data that requires significant expertise to interpret, and analyzing it takes a lot of time.

The 10x platform combines prescriptive algorithms based on chemistry and biology with AI agents that can interpret that data. The team had to do significant work to train the models on the spectrometry data and make their analysis traceable, a key requirement for the tool that will be used to help companies achieve compliance.

Matthew Crawford is a scientist at Rilas Technologies, a firm that performs chemical analysis for other companies – saving clients like biotech startups from having to invest several million dollars in their spectrometry equipment and experts to run it. Crawford has been using the 10x Science platform for several weeks and says it speeds up his work.

Crawford said the model surprised him with its ability to explain its own conclusions, find the right data to analyze on its own, and adapt to testing different types of molecules. While other AI tools he’s tried in the past have over-promised or run into accuracy issues, he says this one makes a logical assumption, something he says involves the deep background expertise of its creators.

“I used a specific protein on it, and it just showed up, from what I named the file, what the protein was probably,” Crawford said. “It then searched online databases to find the sequence of that protein, so I didn’t have to sequence.”

10x executives say they work with many large pharmaceutical companies, as well as academic researchers. The plan is to use this seed funding to hire more developers and continue to refine the model and offer new customers. If they are able to achieve character proteins, Roberts hopes the company will expand to provide a new kind of biological understanding, combining protein structure with other information about cells.

“The profound thing behind what we’re building is actually a new way to describe molecular intelligence,” Roberts said.

For its investors, 10x offers a useful approach to the biotech space that does not depend on the success of a particular drug and winning regulatory approval. If the company works as well as its founders hope, it will become an important tool for drug development, whether or not the products end up succeeding in the market.

“This is a SaaS platform that has to pay pharma, every single month, to go through all these leads,” said Zoe Perret, partner at Initialized. He relies on the deep experience of the founders to protect the company from competitors; there just aren’t that many people who understand these methods and the data they generate.

What the platform can do, Crawford says, is help open up strategies for researchers who could benefit from these methods but don’t have the time or resources to implement them.

“The teams here are trying to make a new drug,” he told TechCrunch. “They just want to get a quick, easy answer to the mass spec, and that opens up a can of worms. This software will help keep that can of worms closed and just get them the answer they need to do the next step in their research.”