“We haven’t yet really figured out how to pay for AI in healthcare,” says Tempus AI CEO Dr. Eric Lefkofsky, a challenge that led to the birth of Tempus, a company that’s impossible to ignore
Lefkofsky’s approach to business is as unconventional as it is relentless. Born in West Bloomfield, Michigan, the youngest of three siblings all University of Michigan graduates, he discovered early on that he had a knack for commerce. As a college freshman, he was already selling carpets to peers and earning $100,000 a year, a success he detailed in a 2012 blog post. “I have never been afraid to act. It is one of my greatest strengths in business. When I have an idea or reach a conclusion, I act on it, immediately and without reservation,” he wrote, a sentiment that would come to define his career.
That fearless drive led him to launch a series of companies that spanned diverse industries. In 1999, Lefkofsky and his law school classmate Brad Keywell moved to Chicago to build Starbelly, an early internet company that was scaled—and sold—for $240 million in less than a year. Over the next decade, each problem Lefkofsky encountered became the seed for a new venture. From InnerWorkings in 2001, designed to manage printing and shipping for books and magazines, to Echo Global Logistics in 2005 and MediaBank in 2006, his entrepreneurial resume grew rapidly. Notably, his involvement in Groupon—seeded with a modest $1 million contribution—ushered him into the Forbes 400 following a meteoric rise and a $13 billion IPO in 2011, despite later struggles.
A Personal Call to Healthcare
Following his wife’s breast cancer diagnosis in 2014 and a subsequent year of conversations with oncologists, Lefkofsky, despite his lack of healthcare experience, founded Tempus in 2015. “I was afraid that it might not make anyone any money,” he recalls, a reflection of his persistent focus on profit. Olufunmilayo Olopade, a respected oncologist at the University of Chicago and Tempus’ founding scientific advisor, recounts a moment when she asked for research funding. “He said, ‘why would I give you $50 million when I can start my own company’ to solve the issue for every university and hospital?” Although Lefkofsky doesn’t remember using those exact words, he agrees that a for-profit model was essential to bring advanced technology and AI to healthcare.
Determined not to repeat past missteps, Lefkofsky slowed the pace he was accustomed to, keeping the company’s focus domestic for much longer than his previous ventures. This deliberate strategy allowed Tempus to build a robust platform centered on oncology before expanding into areas such as psychiatric disorders and cardiology.
From its inception, Tempus has taken a more measured approach than Lefkofsky’s previous ventures. Unlike Groupon, which expanded into 50 countries within three years, Tempus chose to grow slowly and maintain a strong foothold in the United States before branching out. The company recently extended its reach into Japan through a $200 million joint venture with SoftBank, a step that underscores its deliberate expansion strategy. Today, Lefkofsky devotes virtually all of his professional time to Tempus, a contrast to his previous practice of rapidly moving between projects. Even as Tempus waited nine years before going public, a notable difference from his other IPOs that typically took four to five years, Lefkofsky has shown no signs of relenting in his quest for healthcare impact.
A Dual-Pronged Business Model
Tempus operates on two main revenue streams. More than half of its revenue—55% in the first half of 2024—comes from genetic analysis. When a doctor suspects a tumor, a sample is sent to Tempus labs in Chicago, Atlanta, or Raleigh for genomic sequencing. The results are then cross-referenced with a 200-petabyte machine-learning-trained dataset of similar cases. A report follows, detailing the test results, potential treatment options, and links to relevant research. Patient data, stripped of identifying information, is added back into the dataset to refine future analyses.
The remaining 45% of revenue comes from licensing this extensive dataset to research institutions and major pharmaceutical companies. Contracts range from payments of $150,000 to $500,000 per engagement, and on the upper end, multi-year deals with industry giants like AstraZeneca and GlaxoSmithKline—each worth up to $300 million—underscore the value of Tempus’ data assets.
Managing Growth and Internal Culture
When asked about the broader impact of AI in healthcare, particularly in diagnostics, Lefkofsky was emphatic. “Everyone is obviously very focused on generative AI and in particular the benefit of large language models. There’s no area where I think it’s more impactful than healthcare diagnostics. Almost every major decision a doctor makes is made after ordering some kind of laboratory test result. If you can make diagnostics intelligent, you can really route patients to the optimal therapy. Generative AI allows us to use new tools to structure or make sense of disparate information—whether it’s physician progress notes, pathology reports, molecular data, or imaging scans—to ensure patients are always on the optimal therapeutic path. The promise here is enormous, not only in helping patients live better and longer but also in reducing the significant waste in our US healthcare system.”
Even as Tempus navigates intense competition from companies like Foundation Medicine, Flatiron Health, and Guardant Health, Lefkofsky’s management style remains a central pillar of its identity. Described by colleagues as “singularly problem-focused,” he expects frank, concise updates in one-on-one meetings. “If a one-on-one lasted more than five minutes, you knew you were in trouble,” recalls former chief medical officer Gary Palmer, while former chief scientific officer Joel Dudley noted, “If you tell him, ‘Hey, this is going well,’ he’s like, ‘Who cares? Tell me what’s going wrong.’” Such candor has fostered a culture that values honest communication.
Adding to its complex operational landscape, Tempus’ Chicago lab staff unionized in March over concerns about patient care, staffing levels, and workplace safety—the first time an AI company has successfully unionized. Anson Poe, a member of the union bargaining committee, explained, “People felt like they were bringing issues to management and not getting listened to.” Lefkofsky responded, “I don’t have a strong opinion on whether people should unionize, but we’re happy to negotiate.”
In a midtown Manhattan office that contrasts sharply with the company’s Chicago headquarters—with its neatly arranged snake plants and open-floor-plan chatter—Lefkofsky walks into a conference room without closing the door. “I tend to be very singularly problem-focused,” he explains. “I get consumed by it, and then don’t spend time thinking about anything else.”
Strategic Acquisitions
Tempus’ recent acquisition of Ambry Genetics marks a significant step in broadening its offerings. “Ambry is a fantastic addition to our platform. They’re the leader in hereditary screening or looking at inherited risk for patients that might have cancer. And they allow us to round out our platform,” Lefkofsky stated. The deal reinforces Tempus’ commitment to covering the full spectrum of patient care—from risk assessment and therapy selection to monitoring and early detection of disease recurrence.
While still unprofitable on a GAAP basis, Tempus’ trajectory has attracted a host of high-profile investors including Baillie Gifford, Tyro, Google, SoftBank, and even Cathie Wood through ARK funds. Venture capitalist Peter Barris sums up the sentiment: “You have to convince me why I shouldn’t back this guy.” Despite its rapid share-price increase nearly 99% since the start of the year Lefkofsky remains focused on the company’s long-term vision rather than short-term exit strategies. “We’re really more focused on our core vision at hand,” he noted when asked about potential mergers or acquisitions.
On the subject of mergers and acquisitions, especially given his track record, Lefkofsky was clear. “I’m fortunate that I’ve started a bunch of businesses. Four have gone public, and one got sold, but Tempus is at a scale now. Our guidance for this year is $1.23 billion, and we intend to be both cash flow positive and adjusted EBITDA positive. We’re nine years old with the acquisition of Ambry and approaching 4,000 people. We’re in a fortunate place with a business that’s growing quickly and is now going to be making money. So in terms of looking to sell or whatever, we’re really more focused on our core vision. The opportunity is just so big when it comes to bringing AI to healthcare it’s just too big.”
AI-Enabled Capabilities
The company, which went public last year at a $6 billion valuation, reported $600 million in revenue for the year through June, though it remains unprofitable with a net loss of $720 million—more than half of which was tied to an accounting adjustment during the IPO process. Its share price has since increased by 22%, lifting its market cap to $7.6 billion, with Lefkofsky’s stake valued at around $2.5 billion—a sum that now constitutes more than half of his $4.4 billion net worth.
Tempus continues to expand its influence with fresh technology offerings. The launch of Olivia, a personal health concierge app brings patients the ability to consolidate their health data into one secure platform. Patients can sync data from electronic health records, health devices, or manual uploads to create a dynamic timeline that includes lab results, imaging reports, visit summaries, and even DICOM images for direct sharing with care teams. New features such as Smart Profile Summary, an AI-enabled Notetaker, Medical Image Sharing, Health Monitoring, and Direct Data Import are designed to empower patients with clear, actionable insights.
In addition to Olivia, Tempus has enhanced Tempus One, its AI assistant for precision medicine and research. The latest update employs Tempus’ proprietary LLM Agent Infrastructure to process both structured and unstructured data. This upgrade introduces four new capabilities—Patient Query, Patient Timeline, Prior Authorization, and Data Exploration—that aim to help clinicians and researchers better utilize a wealth of data ranging from progress notes and pathology reports to imaging scans. “LLMs now give us the opportunity to derive new insights from unstructured data, which has some of the richest patient data and until now, was extremely difficult to access at scale,” Lefkofsky commented. Dani Castillo, MD, Assistant Clinical Professor at City of Hope, added, “This will be the future for clinicians… Other AI platforms for charts are way more limited and don’t really help generate data with a full report.”
Investors have taken note. Despite a nearly 99% increase in share price since the beginning of the year, many see significant room for expansion. Tempus’ impressive 125% net revenue retention in 2023, along with projected revenue growth of 32% for 2024 and anticipated improvements in adjusted EBITDA, positions the company as a compelling prospect for growth-oriented investors even if the stock carries some risk.
From a founder who once dabbled in diverse fields from apparel and printing to logistics and e-commerce, Tempus AI now stands as what Lefkofsky hopes will be the enduring legacy of his career. “There was never a moment in any company I was ever at where I’m like, this could be as big as Apple,” he reflects.