REFS (Reverse Engineering & Forward Simulation), our patented causal machine learning platform, goes beyond correlation to causation.
This breakthrough platform automates the previously empirical, “trial and error” work of matching interventions to individual patients – precisely, rapidly and across large populations – improving health outcomes, driving down costs, and creating a more effective healthcare system.
REFS-generated models answer questions for each individual patient, such as: How will the patient respond to this treatment? What if we choose one intervention over another?
It infers causal mechanisms between variables – how variable X affects variable Y, if at all, or vice versa. This is based on the mathematics of Bayesian Network Inference and Global Optimization, which led to the awarding of the prestigious “Nobel Prize” of computer science, the Turing Award, to Judea Pearl.
When our customers apply these insights across populations, the optimal future scenarios emerge.
REFS is the only commercially available platform that infers causal mechanisms from patient data at scale from traditional healthcare and emerging data sources.
REFS uses a two-step process – as its name implies, Reverse Engineering and Forward Simulation – to unlock value from the largest, most diverse data streams.
First, REFS reverse engineers causative mechanisms from vast, multi-modal datasets.
In this step, REFS searches for the most likely explanations for the data. Because REFS has no bias and its computational power is highly scalable, it calculates many of the possible combinations of causal relationships that drive the outcomes. Through this process, REFS coalesces around the most probable mechanisms and models that gave rise to the data.
Then, REFS puts these models into action, running “what if?” simulations to determine which treatments and therapeutics will produce the best outcomes for each and every individual in a population.
By identifying the causal relationships between variables, REFS rapidly identifies what works for specific individuals, bringing the promise of precision medicine within reach.