Pittsburgh-based Realyze Intelligence, a company that uses artificial intelligence to connect cancer patients with clinical trials, completed a four-month accelerator cohort that is part of the Biden administration's Cancer Moonshot initiative.
The UPMC Enterprises-backed startup was one of 16 selected to participate in the CancerX Startup Accelerator, which paired startups with mentors, federal partners and professionals across the health care industry to test and validate their products.
"The big part of this was what they call almost matchmaking, how do we connect innovation companies with various players in the space," CEO Aaron Brauser said. "Obviously, there's the health systems and cancer centers that are doing the research and delivering care, but also what was extremely beneficial for us was the other kind of components of that, obviously the pharma companies, the government agencies are very well connected into this program and then other industry components, whether it's investment and also large corporations that serve the area."
One mentor proved pivotal that Brauser said people wouldn't expect — Dell Computers.
"Dell Computers, which people think about as laptops, also does a lot of infrastructure and storage and processing," Brauser said. "They actually were one of our mentors and had great relationships with cancer centers across the countries."
Brauser co-founded Realyze in 2020 alongside Dr. Gilan El Saadawi. In the years since, the AI sector has seen rapid advancements and deployment, but deployment of AI in health care has lagged behind. Brauser attributed this to "a lot of different reasons," one of the biggest being privacy — AI models are trained on data, and getting access to health care data securely is more nuanced than in other sectors due to compliance guidelines. The second major factor is the inherent complexity of the human body.
"The other thing that I don't think people understand in health care is humans aren't inventory on the shelf... you're not just a cancer patient, you might [also] have heart disease," Brauser said. "Every patient is different and the phenotypic representation of you as a person is extremely complex."
This "makes it extremely complex for AI to understand" what the best possible options for a patient are. But Brauser said that the company has benefitted from advancements across the sector.
"We kind of look at it as like a microscope... until someone invented the microscope, there weren't a lot of cell level discoveries that happened," Brauser said. "Once we get this data and the technology there, we're going to be able to find new ways to use the data and have abilities to solve problems we couldn't have even imagined."