--- title: Oncology AI Firm Triomics Raises $22 Million to Expand Across US Cancer Centers description: Oncology AI firm Triomics has raised $22 million led by Battery Ventures to expand its clinical trial matching and chart tools across US cancer centers. author: Darie Nani (Editor-in-Chief) date: 2026-06-03T13:32:18.062Z updated: 2026-06-03T13:32:18.079Z canonical: https://www.sovereignmagazine.com/article/triomics-raises-22m-oncology-ai image: https://cdn.nanimediahouse.com/triomics-founders.webp categories: Artificial Intelligence, Startups, Business content_type: News region: New York publication: Sovereign Magazine about: - type: Organization name: Triomics description: Triomics is a New York oncology AI company founded in 2021 by Sarim Khan and Hrituraj Singh. Its software reads a patient's full cancer record and converts the unstructured notes into structured data, which cancer centers use for clinical trial matching, pre-visit chart preparation and cancer registry abstraction. It is backed by Battery Ventures, Lightspeed, Nexus Venture Partners and Y Combinator. url: https://triomics.com foundingDate: 2021-01-01T00:00:00.000Z industry: Healthcare AI sameAs: - https://triomics.com --- Triomics, a New York startup that develops artificial intelligence for cancer care, has raised $22 million in a Series B round led by Battery Ventures. The round brings its total funding to more than $36 million. Its software reads a patient's full cancer record and turns the notes into structured data. The money will go toward hiring engineers, expanding the teams that install the software in hospitals, and extending its work with companies that run clinical trials. ## Why AI in Oncology Is So Hard A single cancer patient's record can run to hundreds of pages. It mixes clinic notes, pathology and radiology reports, biomarker panels, records from other hospitals, prior treatments and trial eligibility criteria that change constantly. General-purpose AI models tend to get this wrong, because reading across all of it and reaching a conclusion a doctor can defend is harder than handling one document at a time. "Oncology is the hardest place to build AI, yet the most important," says Hrituraj Singh, co-founder and chief technology officer. "Getting a model to reason reliably across thousands of pages of notes, pathology, imaging and evolving trial criteria, and show its work, is what separates a demo from software that clinicians actually use." Triomics says every answer its software produces links back to the source document in the chart, so a clinician can check it. Sarim Khan, the co-founder and chief executive, says hospitals struggle to use the data they already have. ## How Triomics Structures Oncology Data The software uses AI agents to read the full patient record and convert unstructured text into structured, labeled data. That feeds three jobs: matching patients to clinical trials, preparing charts before an appointment, and abstracting oncology data for cancer registries and quality reporting. Hospitals are required to file cancer registry data with state, federal and professional bodies, and the work is done largely by hand. Lee Schwamm, chief digital health officer at Yale New Haven Health System, says the aim is "autonomous chart abstraction of clinical registry quality that can be rapidly reviewed and finalized for reporting by human registrars." Yale uses Triomics for trial matching and is extending it to registry reporting. ## Does AI Clinical Trial Matching Work? The trial-matching tool has the most outside validation. In 2024 the company's model was assessed in a peer-reviewed paper in npj Digital Medicine, part of the Nature portfolio, which found its oncology-specific system roughly matched GPT-4 and qualified physicians at identifying eligible trials, and beat the general-purpose models it was tested against. The company says it has also presented results at ASCO, the main US oncology conference. Triomics also reports gains from customers using its tools: a 40% rise in trial matches, more than 30% more enrollments and a 67% cut in chart review time. The enrollment and screening figures come from a deployment at the Medical College of Wisconsin. These are customer-reported numbers, not findings from the peer-reviewed study, which measured matching accuracy rather than enrollment. Those customers include major US cancer centers: Memorial Sloan Kettering, MD Anderson, Yale Cancer Center and its partner Smilow Cancer Hospital, Mount Sinai's Tisch Cancer Center, and Texas Oncology, one of the country's largest community oncology networks. > "Getting a model to reason reliably across thousands of pages of notes, and show its work, is what separates a demo from software that clinicians actually use." > — Hrituraj Singh, co-founder and CTO, Triomics ## Oncology AI Companies and Market Consolidation The field is consolidating around larger players. Tempus AI, which is publicly traded, bought the trial-matching company Deep 6 AI in 2025. Flatiron Health, owned by Roche, holds a strong position in oncology data, and ConcertAI, IQVIA and McKesson's Ontada work in adjacent areas. Battery's case for backing a smaller competitor is that Triomics runs several of these jobs on one system without separate integrations, which Gleklen calls rare at this stage. "We are live at some of the top cancer centers and demonstrating measurable outcomes, faster enrollment, less manual chart review," says Brandon Gleklen, the Battery principal joining Triomics' board. ## Who Backed Triomics' $22 Million Round Alongside Battery, existing investors [Nexus Venture Partners](https://www.sovereignmagazine.com/article/trifetch-1-9m-ai-specialty-clinics), [Lightspeed](https://www.sovereignmagazine.com/article/coral-healthcare-automation-seed-lightspeed-z47) and Y Combinator joined the round; Triomics was in Y Combinator's 2021 batch. Two strategic backers also took part: Oncology Ventures and Precision Health Informatics, a subsidiary of Texas Oncology. Texas Oncology is both a customer and, through that subsidiary, now an investor in the software it uses. Triomics says its enterprise customer count grew fourfold and its annual recurring revenue grew roughly tenfold over the past year, figures it has not had independently audited. **About Triomics** Triomics is a New York oncology AI company founded in 2021 by Sarim Khan and Hrituraj Singh. Its software reads a patient's full cancer record and converts the unstructured notes into structured data, which cancer centers use for clinical trial matching, pre-visit chart preparation and cancer registry abstraction. It is backed by Battery Ventures, Lightspeed, Nexus Venture Partners and Y Combinator. [Website](https://triomics.com) ## FAQ **Q: What does Triomics do?** Triomics develops oncology-specific AI that reads a cancer patient's full medical record and turns unstructured notes into structured data. Hospitals use it to match patients to clinical trials, prepare charts before appointments, and abstract data for cancer registries and quality reporting. **Q: What is AI clinical trial matching?** It is the use of software to compare a patient's medical history against the eligibility criteria for available trials and flag the ones they qualify for. Done by hand the task is slow and takes up a large share of research coordinators' time, which is why it is a common use for AI in oncology. **Q: Who are Triomics' main competitors?** The largest is Tempus AI, which bought the trial-matching firm Deep 6 AI in 2025. Flatiron Health, owned by Roche, is strong in oncology data, and ConcertAI, IQVIA and McKesson's Ontada operate in adjacent areas. **Q: Has Triomics' technology been clinically validated?** Its trial-matching model was assessed in a peer-reviewed 2024 paper in npj Digital Medicine, part of the Nature portfolio, which found it roughly matched GPT-4 and qualified doctors at identifying eligible trials. The other figures the company cites, such as faster enrollment, are customer-reported results rather than findings from that study.