Every day, thousands of people cross the 520 bridge between Seattle and Bellevue, Wash. But for some of those people, the daily journey is more than just a commute: It is part of a years-long project to create a completely new kind of AI health tool.
Speaking from the health tech stage at the 2018 GeekWire Summit, the leaders of that project described how they have created a tight-knit group of scientists and technologists to take on the challenge, even going so far as to physically integrate the teams.
“It was extremely important to us that we would be working on this thing hand-in-hand, and not with a firewall, a bridge or a lake in between us,” Desney Tan, the general manager of Microsoft Healthcare, said during the panel. He was joined on stage by Adaptive Co-Founder and CEO Chad Robins and journalist Luke Timmerman, founder of the Timmerman Report.
“We really integrated ourselves as a single team,” Tan said. “We’ve got offices in each others’ facilities … The teams function about as tightly as any team can even though we live in two separate organizations.”
Tan said the combined group, which numbers between 40 and 50 employees, spends a lot of time traveling between Microsoft’s headquarters in Redmond, Wash., and the Adaptive office in Seattle.
But closing the physical distance is relatively easy. It’s quite another challenge to integrate a deeply scientific company like Adaptive, which focuses on the intricate details of human biology, with a tech giant whose expertise lies in advanced artificial intelligence.
Robins said Microsoft has more than risen to that challenge.
“They’ve been extremely impressive on the biology,” he said. Adaptive’s technology sequences the unique biological markers on immune cells. The company will work with Microsoft to train an AI tool to read that data and detect any diseases the immune system is fighting, effectively creating a tool that can diagnose any disease from a simple blood draw.
Last month, the FDA cleared Adaptive’s clonoSEQ device, which uses the same immune sequencing technology to detect residual cancer cells that survive treatment. The clearance could be good news for future technologies that take the same approach.
Tan said the smooth sailing was partially due to the team the companies were able to assemble.
“We built a team that has been around for a while, and Jonathan Carlson — who runs a lot of the technical part of the project — was a computation immunologist who had been at Microsoft for a little over 10 years working on HIV vaccines,” Tan said. They also pulled talent from Microsoft’s genomics group and hired additional computational biologists to work on the project.
“There was just enough overlap that we could get in the room together and teach each other stuff very, very quickly early on,” Tan said.
On Adaptive’s side, Robins said the company actually had offers on the table from a number of tech companies that wanted to work on the project. He said there were several reasons the company eventually chose Microsoft.
“First and foremost is the people,” Robins said. “The people and the level of commitment from senior leadership.” He said that he and his brother and co-founder, Harlan Robinson, were able to present their work and speak with Bill Gates and Satya Nadella before they decided to partner with Microsoft.
“It was clearly a cut above the other teams that we talked to, in terms of the level of knowledge about the biology and then the thoughts on how to apply the machine learning to the problem,” he said.
And why did Microsoft see Adaptive as a good partner?
“These guys are going to change the world, and we’re thrilled to be a part of it,” Tan said.
Watch Tan and Robins’ full talk at the top of this story and check out all the coverage from the 2018 GeekWire Summit, including more coverage from the Health Tech stage.