oFM: The Oncology Foundation Model
A holistic, patient-level foundation model built specifically for oncology
Overview
Built on real-world data to predict real-world outcomes
oFM is powered by one of the largest, de-identified multimodal datasets in the world, including longitudinal clinical records, molecular profiles, and pathology images, to map how a patient’s disease changes over time.
Models require appropriate validation and market authorization, as required by applicable law, prior to clinical deployment.
The Tempus difference
A single, interconnected oncology ecosystem built to move your pipeline forward.
Multimodal RWD
High-fidelity, de-identified datasets across clinical, molecular, and pathology modalities
Patient trajectories that track longitudinal patient histories and disease trajectories over time
Proprietary Foundation Models
State-of-the-art, multimodal foundation model portfolio trained on vast, proprietary data to capture complex disease context
Integrated, full stack solutions to accelerate your development timelines
Scientific and AI Expertise
Collaborate with experienced scientists and AI experts to rapidly test and validate hypotheses
Deep domain expertise translating scientific hypotheses and asset priorities into AI insights and products with real pipeline impact
Connected Network
Vast clinical footprint capturing diverse, representative real-world data at the source
Seamless distribution channel to deliver AI insights back to physicians at the point of care
How it’s used
We trained oFM on massive datasets to create a holistic understanding of disease across multiple modalities and scales, from the molecular level to the complete patient trajectory. This powerful starting point accelerates AI development, enabling us to rapidly adapt and fine-tune models for your specific use case, requiring significantly less additional data and compute.
Research
Decode complex disease biology
Identify novel signatures of response and resistance: Discover complex, multi-omic phenotypes and patient profiles that correlate directly with therapeutic benefit, and map the underlying biological factors driving patient sensitivity or resistance to targeted therapies.
Interrogate mechanisms of action: Identify novel, high-dimensional phenotypic subtypes using multimodal data, helping to define the relevant mechanistically matched population.
Characterize spatial biology and the tumor microenvironment: Analyze spatial features and cellular structures from pathology images to see how they influence disease progression.
Map biomarker expression: Evaluate how genomic markers and phenotypic expression, including tissue morphology, interact to shape holistic disease behavior.
Clinical Development
De-risk pipeline decisions and increase trial success
Predict treatment benefit and risk: Predict payload sensitivity, therapeutic benefit, and progression risk to optimize pipeline prioritization.
Simulate patient-level Digital Twins: Predict specific patient outcomes on standard of care regimens, and their benefit from experimental regimens, based on their individual biology and clinical history.
Refine clinical trial strategy and design: Utilize retrospective trial analysis to select the relevant indication and patient populations, and inform endpoint strategy to maximize trial success.
Develop digital biomarkers: Build predictive and prognostic biomarkers from high-powered, multimodal data and identify the subpopulation who benefits from experimental assets.
Stratify patient prognosis: Evaluate baseline progression risks to ensure balanced cohorts and precise risk-adjustment during trial design.
Commercialization
Maximize patient access and label expansion
Support label expansion: Leverage robust predictive insights to identify new indications or lines of therapy where existing assets will deliver maximum clinical benefit.
Deploy AI-enabled diagnostics in the clinic: Develop, validate, and bring to market AI-enabled Companion Diagnostics and AI Software as a Medical Device through centralized and decentralized channels.
Model specifications
Partnership Framework
We offer flexible frameworks to align our AI capabilities with your needs
Partner with us at any stage, from proof of concept to commercialization, on a customized project tailored to your unique needs.
Foundation Model Licensing
Access foundation model embeddings for further development in your own environment
Control over training data and downstream AI development
Leverage your own compute resources
Combine access across multiple teams
Pre-built AI applications
Leverage off-the-shelf AI applications for rapid insights, across our data or yours
Tools for rapid insights at the scale you need
Pan-cancer digital biomarker screening
TME characterization and spatial biology
Featured Resources
Webinar
Article
Press Release
This is AI-enabled precision medicine
This is the future of healthcare.