For an investment discussion, please email Dr Mike Tremblay at mike_tremblay@skythunder.net
What underlies the predictors?
The predictors use machine learning to support defined clinical decision making by clinicians. The core intellectual property lies in the methodology for translating complex clinical problems into usable predictive applications, including model design, feature selection, patient-specific baselines, risk estimation, explainability and decision-focused outputs.
What has been built?
The portfolio includes applications at different stages of development. Some are interactive simulations that define functional and user requirements. Others are close to minimum viable product stage, with elements already implemented in Python.
What has been tested?
Most applications have been developed and tested using synthetic patients and simulated clinical scenarios. This has allowed the model logic, workflow, outputs and response to changing patient conditions to be examined before access to real-world clinical data.
What evidence remains to be generated?
The principal evidence requirements are:
- predictive performance, calibration and accuracy;
- value for money;
- adoption within routine clinical workflow;
- usability and clinician acceptance;
- effect on clinical outcomes, costs and decision quality.
The exact evidence programme will depend on the intended use, setting and regulatory status of each application.
Who pays?
The users are clinicians and healthcare organisations. The expected purchasers are health systems, hospitals, care providers and other organisations that can realise value through:
- improved patient outcomes;
- reduced avoidable costs;
- fewer clinical errors;
- lower litigation exposure;
- more efficient clinical workflow;
- better use of scarce clinical expertise.
What is the commercial offer?
Applications may be acquired through:
- outright purchase;
- licensing;
- royalty arrangements;
- co-development;
- investment in further development.
Commercial terms would be negotiated for each application.
Cassis also welcomes investment in additional clinical predictors and in the further development of reusable foundation tools that support multiple applications.