What is the Triomics origin story? What’s interesting about this business model?
My cofounder Hrituraj and I have known each other since college. My focus was biotech and his was AI, but we both wanted to be entrepreneurial from our early days in college. When I was a biotech researcher at MIT after college, I saw challenges from our hospital collaborators in terms of completing time- and labor-intensive workflows. I realized Raj and I had complementary backgrounds and what seemed like the right cofounder skills to tackle a problem like this, so I called him.
He was an Adobe AI researcher at the time, and we realized that recent advances in AI were perfectly suited to address hospital staffs’ pain point of having to manually review free-text (“unstructured”) entries in patient charts. We therefore decided to partner to build solutions leveraging the advances in the field of generative AI and LLMs to help hospital staff process vast amounts of free-text patient data for key operational tasks, such as matching patients to clinical trials, or abstracting patients’ records for cancer registries.
What kind of challenges are you solving? What is the traditional approach and how are you improving on that?
Within healthcare and specifically within oncology, a vast amount of key data is “unstructured”. This makes searching through patient records (efficiently) difficult if not impossible, which in turn hampers many tasks related to clinical research and care. The current gold standard approach is manual chart review, where trained staff combs through the entire longitudinal patient record manually to identify relevant data points.
Our technology stack — an oncology-focused large language model (OncoLLM™), paired with custom software we build for each specific use case — allows for this process to be completed at scale in a fraction of the time. Harmony is our solution for EHR data curation for registry or research needs, while Prism is our solution for patient-trial matching. Finally, we engage in ongoing conversations with our partners about our future product roadmap.
How do you describe your typical users? Who are your target customers?
Our end users are mainly hospital staff, specifically clinical research nurses / coordinators, registrars and other stakeholders focused on improving oncology research and care within their institutions. We partner with large academic medical centers, health systems, and cancer centers.
Can you give me some use cases of how people can / have benefited from Triomics?
In a fraction of the time vs current workflows:
Clinical research coordinators can complete patient-trial matching. Patients ultimately benefit as they are more quickly enrolled in potentially life-altering trials, as do trial sponsors (often pharmaceutical companies), who are able to enroll their studies more completely and quickly, enabling them to bring successful candidate drugs to market faster.
EHR-to-EDC integration (connecting electronic health records to the electronic data capture systems that store clinical trial data): see above beneficiaries.
Registrars can abstract necessary data for tumor registries. Etc.
More generally, our blog discusses in detail the work overload that providers face daily + the associated stress and job dissatisfaction.
How much total funding have you got so far?
We have raised approximately $15 million in VC funding from leading Silicon Valley investors, including Y Combinator, Lightspeed, General Catalyst, and Nexus Venture Partners.
What stage of revenue are you in? How do you/will you monetize?
We have signed our first few deals at the start of 2024, and are targeting 10-15 partner institutions by the end of the year. The commercial structure for each partner institution is unique and tailored specifically to the use case(s) we are helping them to solve. When possible, we are passing through costs directly to pharmaceutical sponsors of clinical trials, to reduce the financial burden on our provider partners, as well as entering into performance-based agreements to ensure our partners see return on their ROI before having to pay anything out of pocket.
Which companies do you consider your closest competitors? How do you differentiate?
We do not envision ourselves competing directly with foundational model companies like OpenAI, Anthropic, Google or Microsoft. We differentiate ourselves by specializing specifically around oncology and developing software that is tailored to complex use cases and specific end users. While there are other vendors trying to solve some of these use cases, such as patient matching, these companies are not GenAI native, but rather rely upon utilizing / modifying legacy technologies without the scale benefit or step-function ROI the industry is asking for.
In the future, how do you plan to expand your platform?
We believe there are an endless amount of use cases that our core technology (OncoLLM) could power. We look forward to pioneering GenAI usage in oncology research and care alongside our provider partners throughout this year. Later this year, we expect to build products and solutions specifically for pharmaceutical and life sciences companies, and their unique pain points and value-drivers.
Tracy Williams
Triomics automates oncology workflows with GenAI powered platform, raises $15M
