Bryan Kyritz

Eric Lefkofsky co-founded Groupon. In 2014 his wife, Liz, was diagnosed with breast cancer, and he was struck by how little data shaped her care. A year later he founded Tempus. Today it reported $411.5 million in revenue for a single quarter. The third quarter of 2026: revenue up 23%, oncology test volume up 32%, and about 12,000 tests for minimal residual disease, which look for traces of cancer in the blood after treatment, up 300%. Adjusted EBITDA went from $1.5 million a year ago to $25.6 million. Full-year guidance is now $1.60 to $1.605 billion. What I find most interesting is how the model works. Every tumor Tempus sequences helps a patient today and also adds to a library that now holds 600 petabytes of clinical and molecular data. Drug companies license that data for discovery and development, with more than $550 million in data bookings so far this year, including a new deal with Bristol Myers Squibb. Each test makes the next diagnosis, and the next drug, a little smarter. I also came to medicine from telecom and the internet, after Viatel, Jazztel and Fon. What surprised me most is how long everything takes. Tempus spent years losing money while it built that library, before it became an advantage. Eric and his team stuck with it, and patients are better off for it. At Certuma we're learning the same lesson in medical AI. Real clinical cases are the foundation, and you build it one patient at a time. The most valuable health data is the kind collected while helping someone.