Pytho LLC
2 min readSep 5, 2019



NSF’s Decision, Risk and Management Sciences Program Awards a Research Grant Teaming Johns Hopkins University with Pytho LLC

Brooklyn, NY, September 9, 2019 Pytho LLC has received National Science Foundation (NSF) funding of $327,360 to test its Human Forest method (patent pending) of predicting which cancer drugs will succeed in clinical trials. The project, titled “Human Forests versus Random Forest Models in Prediction,” pits Pytho’s human-centric prediction method against artificial intelligence models developed by researchers at Johns Hopkins University, which will serve as the prime administrator of the award.

Pytho co-founders Pavel Atanasov and Regina Joseph have decades of experience in prediction and collective intelligence. As members of prior and current U.S. government-funded research teams breaking ground in the science of decision-making and quantified forecasting, their work has received international attention. Joseph is a recognized Superforecaster and Atanasov a noted scholar in distilling the wisdom of crowds. Their Human Forest method poses a potential breakthrough in outperforming computer predictions of new drug success by harnessing people’s collective brain power.

“The power of human prediction is often overlooked due to the hype around AI, but our experience has taught us that human collective intelligence — when harnessed with the right tools — can dramatically improve on the performance of pure machine models,” says Regina Joseph, Pytho co-founder and Co-Principal Investigator on the project.

“We developed our Human Forest method as a way to help people avoid bias and combine their unique insights for solving complex problems and we’re excited to test this against state-of-the-art machine-learning models on tasks like clinical trial prediction,” says Pavel Atanasov, Pytho co-founder, proposal lead and Co-Principal Investigator on the project.

“The great thing about collaborating with an innovative startup like Pytho is that the boundaries of academic research are pushed beyond what we thought would be possible or applicable in the real world. We were primarily focusing on data and modeling methods to predict success in clinical trials, but Pytho has pushed us into thinking about another aspect, which is understanding the value of human expertise in prediction.” notes Sauleh Siddiqui, Assistant Professor at Johns Hopkins University’s Whiting School of Engineering and Principal Investigator on the project.

For all inquiries, contact

NSF’s Decision, Risk and Management Sciences program supports scientific research directed at increasing the understanding and effectiveness of decision making by individuals, groups, organizations, and society.