Clinical trials are notoriously slow, but what if the process could be sped up, improve patient outcomes and save everyone time and money?
Unlearn is trying to do just that by creating digital twins on its machine learning platform.
The San Francisco startup announced a $50 million Series B round on Tuesday that was led by Insight Partners and also included Radical Ventures, 8VC, DCVC, DCVC Bio and Mubadala Capital Ventures. This brings the company's total funding to just under $70 million.
It was co-founded in 2017 by CEO Charles Fisher, head of machine learning Aaron Smith and head of data science Jon Walsh.
Fisher also has a background in biophysics. He received a PhD in biophysics from Harvard and was working towards another post-doctoral degree — and a career as a professor — at the École normale supérieure in Paris but then a friend recruited him for a job at Pfizer.
In 2016, Fisher moved out to San Francisco where he met Smith and Walsh while they were all working as machine learning researchers at San Francisco-based VR startup Leap Motion.
Shortly thereafter, they all left Leap Motion to start Unlearn.
The company takes historical clinical trial data, analyzes it and creates digital twins of real patients to help pharmaceutical companies find trial participants more quickly.
"Over the past couple of years, everyone saw how important it is for these things to be run not only quickly but that what you get out of them too is the right answer," Fisher told me. "Sometimes people think that those two things are at odds. That in order to get a fast answer, you have to maybe lower the bar of evidence. One of the biggest things that we think about at Unlearn, is that's really not true. That new technologies allow you to have both."
By cutting down on the amount of time it takes to enroll patients, ideally by 20%, Unlearn saves clinical trials money and reduces the chances of a participant not being well-suited for a particular trial.
It calls its technology prognostic covariate adjustment, or PROCOVA.
Unlearn currently partners with a few pharmaceutical companies including Merck KGaA, Darmstadt in Germany. In March the European Medicines Agency, a regulator similar to the FDA, published a draft opinion that Unlearn's methodology can be used in phase 2 and phase 3 clinical trials.
The company has around 40 employees, mostly based in the Bay Area, and this new funding will help them expand its headcount across all operations including machine learning, biostatistics and finance.
Fisher also wants to grow the scope of the company's technology.
"One of the big uses for this funding round is to expand the number of indications that we're working in," Fisher said.
They have tested their digital twin technology on clinical trials for Alzheimer's disease and the partnership with Merck in Germany focused on immunology, Fisher told me.
