Faro Raises $37.3M Series B to Put AI Agents Inside Clinical Development

Most of the money flowing into AI drug discovery has gone to molecule design. Isomorphic Labs raised $2.1 billion, Chai Discovery $400 million. The problem they attack is finding candidates faster and cheaper. But the bulk of a drug’s development timeline sits in what comes after: the clinical trials. Even with a molecule in hand, the road from first-in-human studies to approval runs for years. Far fewer companies are working on that stretch.

Faro, based in San Diego, has raised $37.3 million in a Series B co-led by Merck Global Health Innovation Fund and S32, with every existing investor participating and Ankona Capital joining as a new backer. The company says six of the ten largest pharmaceutical companies in the world now use its platform.

Structuring Intent, Not Documents

Faro started with the clinical protocol because it sits at the center of so many downstream development activities. But making protocol authoring faster was never the goal.

Clinical development runs on thousands of interconnected scientific, medical, regulatory and operational decisions. Change one part of a study and it ripples through eligibility criteria, site operations, data collection, statistical analysis, regulatory documentation and every system downstream. For AI to take a meaningful part in that, the company argues, access to documents is not enough. It has to understand the intent behind a development program.

So Faro spent years building proprietary clinical development data models that represent a study’s concepts, relationships, constraints and intent in a structured, machine-readable form. On top of that, software and AI agents can reason across the whole of clinical development instead of treating each document or task as an isolated interaction. Documents remain important outputs, the company says — they are simply not the underlying system of record.

Faro frames the difference this way: an agent that drafts a section of a protocol is useful. An agent that understands why an endpoint was selected, how that decision affects the schedule of activities, what data must be collected, which downstream systems need configuring and which regulatory and operational constraints have to be preserved is something else entirely.

Customer ambitions are widening too. Teams are thinking across study design, protocol authoring, study build, execution, data flows, review and the downstream work needed to move a program forward. The value of AI in clinical development, the company expects, will come from reasoning over connected decisions and taking action while maintaining the oversight a regulated environment demands. The new capital goes toward expanding Faro’s agentic AI capabilities and supporting customers deploying agents across their development organizations. The same foundation could extend well beyond a protocol or any single stage, the company believes, because biopharma is full of workflows where scientific intent, structured data, operational decisions and regulatory requirements all have to stay connected.

Faro was founded in 2019 by Scott Chetham and Ross A. Jaffe, with Chetham serving as co-founder and CEO. It began with a $15 million Series A in June 2021 and added $20 million in February 2023 in a round led by General Catalyst, joined by Section 32, Polaris Partners, Zetta Ventures and Northpond Ventures. S32, co-leading this round, is that same Section 32 under a new name. Including this Series B, total funding reaches roughly $76 million. PTX Capital also participated as an existing investor.

The detail worth noting is Merck Global Health Innovation Fund coming in as a new co-lead. A pharmaceutical company’s strategic fund taking the lead means capital arriving directly from the side that would become the customer.

Who Else Is Attacking Clinical Trials With AI

The field is filling in. Lindus Health raised a $55 million Series B for an approach that runs trials end to end on a client’s behalf — where Faro builds tools for teams inside pharma companies, Lindus operates the trial itself. Scoop focuses on compressing regulatory submission paperwork from months to days and came out of Y Combinator last year. Strand AI emerged earlier this year with an approach to filling gaps in clinical trial data using AI.

What separates Faro is that the structured data model sits underneath rather than the document sitting on top. Plenty of tools will write a protocol quickly. Making the intent behind a study design machine-readable is what allows several workflows to be stacked on the same foundation.