Blood Tests Will Catch Returning Cancer 12 Months Early | Tempus

Eric Lefkofsky (0:00)

You're able to, from blood, detect cancer coming back, like 6 or 12 months before a typical scan would detect it. So in the name, in cancer, the name of the game is you have to kind of treat people quick, otherwise time is not your friend. And so if we treat a patient, and we think they're doing well, and then all of a sudden it takes 9 months or 12 months for us to do a scan and find out the cancer has come back, that can often be too late to pivot to some other treatment, we could have lost the whole year.

If we can detect that in 8 weeks, or 10 weeks, or 12 weeks, we're now putting them on a different therapy right away and trying to get ahead of it. So that's a really cool innovation in this space that I think will become pervasive.

SPEAKER_2 (0:44)

Hello everyone and welcome back to MTS. Today we are joined by Eric Lefkofsky, the founder and CEO of Tempus AI and the former co-founder of Groupon. So today we're gonna talk a little bit about how AI is changing personalized health care and all the latest when it came to Tempus Q3 release. So Eric, welcome to MTS.

Eric Lefkofsky (1:06)

Thanks for having me.

SPEAKER_2 (1:07)

Thank you for joining. I know Q3 say was a very strong quarter for Tempus. Can you share a little bit more of those results?

Eric Lefkofsky (1:14)

Yeah, I mean, it was another record quarter for us. Our revenue was up, I think 23% led by really strong growth in our main cancer testing business, where units increased like 32%.

Our data and AI applications business was up 33%. So our two largest businesses are both growing in the low to mid 30s. And we just have a ton of tailwind. We're in a really good spot. Our sequencing business is just experiencing a lot of strength. And as AI becomes more pervasive in treating cancer patients and in health care more generally, we just have a ton of demand from Big Pharma and Biotech for our data and our modeling capabilities. And so that business has just been exploding with, I think we've signed three out of $100 million plus deals this year alone with Merck and Biotech and BMS. And so it's just really strong. Yeah.

SPEAKER_2 (2:17)

I want to talk a little bit about where the landscape is right now, because you have been working on Tempus for over a decade. I want to understand what's different about 2026, whether that's model capabilities, how the wider landscape is talking about AI and health care, and what's possible, what's unique today, this year, that wasn't true a decade ago.

Eric Lefkofsky (2:40)

Well, I think for us, we have long believed that by aggregating large amounts of rich multimodal data, especially for cancer patients, where you have rich molecular data and rich clinical data, and where the clinical data is updating regularly so you can follow these patients longitudinally over time, we've long thought that this was the kind of rich data set you would need to really figure out how to do two things, ensure that every patient is on the right therapeutic path, and improve research so that you make drugs and therapies way more effectively. Up until the last few years, we had enormous amounts of data, but really lacked a bunch of the tools to interrogate that data at scale. We have 600 petabytes of data, so it's a huge amount of data, and it grew by 50 petabytes last quarter alone.

You really need tools that allow you to interrogate structured clinical data, unstructured data like pathology reports, submission progress notes, images, pathology slides, radiology scans, rich molecular data. Those tools all have gotten to real maturity and scale over the last few years with these most recent models that have come up from not just the frontier modeling folks with RoboScript versions. And so we now can do things that were just like impossible a few years ago. So it's become pervasive throughout our entire product portfolio.

SPEAKER_2 (4:09)

Yeah. A big time when I talk about AI's impact and how that actually affects the everyday person. The thing that resonates the most is when it comes to healthcare. It's very tangible. You can grasp it. You understand what that actually means.

I wanted to talk a little bit about Tempus' origin story. This is a beautiful origin story and very personal. Can you share a little bit more there?

Eric Lefkofsky (4:33)

In my case, I started Tempus about 11 years ago, after my wife was diagnosed with breast cancer, just perplexed and amazed how little technology was a part of her care.