Triomics Secures $15M to Automate Oncology Workflows with GenAI

Triomics co-founders: Sarim Khan and Hrituraj Singh

What You Should Know: 

– Triomics, a company leveraging generative AI to revolutionize cancer care, today announced securing $15 million in funding led by prominent Silicon Valley firms like Lightspeed, Nexus Venture Partners, General Catalyst, and Y Combinator. 

– Triomics partners with leading academic cancer centers and researchers to develop generative AI performance and safety benchmarks and best practices. Triomics partners include the Collaboration for Oncology-focused LLM Training (COLT), a consortium of leaders from a dozen NCI-designated cancer centers, and the Cancer Informatics for Cancer Centers (CI4CC) Society.

Transforming Oncology Workflows with GenAI

Currently, oncology staff grapple with manually searching massive amounts of patient data to identify suitable clinical trials and treatment pathways. This time-consuming process often leads to:

Delays in patient care: Manual data analysis can take hours per patient, potentially causing patients to miss out on critical clinical trials or biomarker-driven treatments. Backlogs and inefficiencies: Healthcare systems often struggle with backlogs in oncology workflows, impacting overall quality of care. Provider burnout: The overwhelming workload can contribute to provider dissatisfaction and burnout.

Triomics’ Solution: OncoLLM™ and AI-powered Software

Triomics co-founders, Sarim Khan (CEO) and Hrituraj Singh (CTO), recognized the limitations of existing software that only analyzes structured data. They identified the potential of generative AI to unlock the vast potential of unstructured medical data, like doctor’s notes, which comprises roughly 80% of patient information.

OncoLLM™: Unlocking the Power of Unstructured Data

Triomics developed OncoLLM™, a generative AI model specifically trained on oncology data. In collaboration with researchers from the Medical College of Wisconsin, Triomics demonstrated that OncoLLM™ can:

Identify eligible clinical trial patients in minutes: A process that would take qualified nurses days or weeks is significantly accelerated, potentially improving patient access to critical therapies. Extract data from unstructured notes with high accuracy: OncoLLM™ achieves similar or better accuracy compared to leading models like GPT-4 or Claude, at a significantly lower cost (40 times cheaper). Outperform other retrieval models: Triomics’ information retrieval engine has been shown to be 1.5-2 times more effective than other state-of-the-art models.

Triomics Software Integrates AI with Existing Systems

OncoLLM™ powers Triomics’ software suite, which seamlessly integrates with electronic health record (EHR) systems used by hospitals. This allows for functionalities such as:

Triomics Prism: Automates patient-trial matching by prescreening oncology patients for relevant clinical trials based on upcoming appointments. Triomics Harmony: Curates and analyzes EHR data to support quality reporting, cohort analysis, and personalized cancer care (precision oncology).

Triomics plans to publish further research on OncoLLM™’s efficacy across diverse patient populations and healthcare settings.