Somewhere in the labyrinth of biomedical research, a common frustration persists: scientists who can decode the human genome still struggle to share their findings across institutional firewalls. Enter Manifold, a Newton-based company that believes it has cracked the code—or at least part of it.
The startup announced in mid-December that it had raised $18 million in Series B funding, led by Reach Capital with backing from SilverArc Capital, Industry Ventures, and returning investors TQ Ventures and Calibrate Ventures. The round brings Manifold's total funding to $40 million, a relatively modest war chest for a company aiming to overhaul how the $2 trillion life sciences industry handles data.
What makes Manifold's pitch distinct is less about the scale of its ambition than the specificity of its target. The company isn't trying to build general-purpose AI. Instead, it describes its product as "vertical AI"—tools designed exclusively for the peculiar demands of regulated healthcare environments where data governance isn't optional.
The Infrastructure Play
Founded in 2016 by Vinay Seth Mohta, Vivek Mohta, and Sourav Dey, Manifold has spent nearly a decade constructing what it calls an "Agent OS" for biomedical research. Think of it as middleware, sitting between raw genomic data and the researchers trying to make sense of it.
The platform now serves tens of thousands of users across hundreds of organizations worldwide, according to the company. (Exact figures remain closely held.) It ingests multimodal biomedical data—genomics, clinical records, imaging files—and layers on governance controls that accommodate healthcare's Byzantine regulatory landscape.
Two flagship features illustrate the approach: an AI Cohort Explorer that helps researchers identify patient populations matching specific criteria, and an AI Analysis Agent designed to automate chunks of the research workflow that typically require manual tinkering. Whether these tools genuinely save time at scale remains an open question, though early adopters seem convinced enough to keep using them.
Perhaps more notable is Manifold's stewardship of Terra, the genomics research platform originally built by the Broad Institute, Microsoft, and Verily. The company has been quietly modernizing Terra on AWS infrastructure, collaborating with the Broad Institute's Data Sciences Platform team on what they announced in January as a next-generation, AI-enabled research environment. It's an unusual arrangement—a startup operating a platform birthed by research giants—but one that appears to be working.
Partnerships That Signal Intent
If the Series B validates Manifold's business model, the partnerships it has struck in recent months suggest broader industry acceptance.
Last October, Foundation Medicine—a heavyweight in genomic profiling—launched "FoundationInsights powered by Manifold," bringing AI-enabled analytics to its database of over 800,000 patient genomic profiles. The tool went live immediately for biopharma partners, with a version tailored for healthcare providers expected sometime in 2026. That kind of collaboration doesn't materialize unless the underlying technology passes rigorous vetting.
Manifold has also woven itself into the ecosystems of Anthropic and AWS. The Anthropic tie-up integrates Claude, the AI assistant, into governed scientific settings—an acknowledgment that generative AI and strict compliance protocols can, in fact, coexist. Meanwhile, AWS's health-focused tools, including HealthOmics and Bedrock, now operate within Manifold's platform.
On the academic side, institutions like the University of Virginia Comprehensive Cancer Center, Indiana University's Simon Comprehensive Cancer Center, Emory's Winship Cancer Institute, and Morehouse School of Medicine have all adopted the platform. The Morehouse collaboration, announced in September 2024, focuses on integrating social determinants of health with genomic data to better understand cancer disparities—an ambitious and politically resonant research goal.
A Crowded but Fragmented Market
Manifold isn't alone in chasing this opportunity. DNAnexus, which pulled in $200 million in 2022, operates in similar territory. So does Lifebit, which runs federated data intelligence platforms for national genomics initiatives. The market is crowded but fragmented, with no clear dominant player.
Manifold's differentiator, at least according to its pitch, lies in marrying governance controls with generative AI. In clinical and pharmaceutical research, where a single data breach can torpedo years of work, that combination isn't just useful—it's existential.
The company now employs somewhere between 51 and 200 people, per LinkedIn data. It secured its $15 million Series A from TQ Ventures in April 2024, meaning this Series B arrived in less than two years. That's a relatively brisk fundraising cycle, suggesting either strong traction or investor enthusiasm. Possibly both.
What Comes Next
Manifold plans to deploy the fresh capital toward deepening its collaborations with the Broad Institute, AWS, and Anthropic, while continuing to modernize Terra. Translated: more infrastructure work, more integrations, more attempts to become indispensable.
The broader question is whether Manifold can scale without losing the precision that makes its platform appealing in the first place. Vertical AI is valuable precisely because it's not general-purpose. But niche markets have ceilings, and those ceilings can arrive sooner than founders expect.
Still, in a field where the ability to query complex datasets in natural language—without triggering compliance alarms—could shave months off drug development timelines, there's real money to be made. Whether Manifold can capture enough of it to justify its valuation remains the story investors are betting on.
For now, the company is doing what most well-funded startups do: hiring, building, and trying to convince the market that its version of the future is the one worth backing.
