• Healthcare-focused AI startups are raising billions to help improve the US system.
• AI can help streamline clinical documentation, drug research, and medical billing.
• This article is part of "Trends in Healthcare," a series about the innovations and industry leaders shaping patient care.

The founder of Suki, a startup that uses artificial intelligence to automate healthcare documents, raised $70 million in funding from investors in a Series D round that was disclosed this past fall.

He said it really didn't take that much persuading: With an epidemic of stressed- and burned-out physicians, there was an obvious need for their AI software, he added.

"Most of the investor conversations over the last year and a half have been, 'Well, it looks like the market is here,'" said Punit Singh Soni, Suki's founder. "Are you going to be the winner or not?"

Suki sells an AI-powered assistant that takes notes during a conversation between patients and clinicians. The notes can be reviewed by the doctor and submitted as an electronic health record. This saves time on administrative tasks and allows physicians more time to take care of patients, a resource that's becoming increasingly limited among healthcare professionals.

Surveys have consistently found that doctors and other medical workers are burned out from working in an often overloaded, convoluted, and inefficient system. The US spent $4.8 trillion on healthcare in 2023, according to a January report from the Peter G. Peterson Foundation. The US also spends more per person than nearly all other developed nations, according to a report by the Organization for Economic Co-operation and Development. Despite this, health outcomes were poorer, with Americans facing a lower life expectancy, higher rates of treatable and preventable excess deaths, and less efficient healthcare systems.

Cash-strapped hospitals and private practices have lagged behind the financial-services and telecommunications industries in applying newer technologies, but the healthcare industry is increasingly considering artificial intelligence as it contends with high labor costs and a lot of opportunities to automate routine tasks. The pandemic exacerbated these challenges with staffing shortages as overworked doctors and nurses quit the profession.

To make healthcare more efficient, AI startups like Suki, Zephyr AI, and Tennr have raised millions with vast promises, including making repetitive tasks like billing and note-taking easier, improving the accuracy of clinical diagnosis, and identifying the right patient population for emerging treatments.

But the challenges are vast. The healthcare industry's budget allocations for generative AI are trailing those of many other core industries, such as energy and materials, consumer goods, and retail. Clinical diagnosis will continue to require a human in the loop, so the process can't be fully automated. The healthcare industry is highly regulated, and quite often, venture capitalists will wait for clarity on laws from the federal government before aggressively pushing AI tech advancements forward.

A $370 billion bet on boosting the healthcare sector's productivity

The consulting firm McKinsey estimates that generative AI can boost productivity for the healthcare, pharmaceuticals, and medical-products industries by as much as $370 billion by accelerating drug research, making clinical documentation easier, speeding up medical billing, and helping doctors make diagnoses.

Some big funding rounds announced in 2024 highlight the diverse use cases for AI in the healthcare sector. They include $150 million raised by the clinical-documentation AI startup Abridge in February, the drug-discovery AI startup Xaira Therapeutics bringing in $1 billion before its launch in April, Atropos Health's $33 million Series B in May to help doctors analyze real-world evidence with generative AI, and the medical-billing-automation provider Candid Health raising $29 million in September.

Parth Desai, a partner at Flare Capital Partners, has steered investments into healthcare startups such as Photon Health and SmarterDx. He said that healthcare organizations had been dedicating more money to bolster their AI strategies, beginning in late 2022 and accelerating through 2024. That's boosting demand for the tools these startups are developing. There's also less pressure to immediately prove a return on investment, which budget-conscious health systems have closely monitored in the past when allocating dollars for technology.

"The thing that we're really studying before making an investment decision is: Do budgets exist today to pay for this technology?" Desai told Business Insider. "Or are they going to exist in a large-enough fashion in the next five to 10 years to support this technology?"

Candid Health and Akasa aim to cut costs and automate medical billing

One area of particular promise has been medical billing, which could benefit from large language model automation. An LLM could, for example, analyze a large volume of claims in a client's system and accurately match them with insurers' unique billing codes, a process required for repayment to a physician for their services. Hospitals have traditionally relied on human medical coders to hunt down reimbursements from insurers.

"The software used to do billing was built a long time ago and basically wasn't kept up to date," Nick Perry, a cofounder and the CEO of Candid Health, said.

Malinka Walaliyadde, the CEO of Akasa — another medical-billing-focused AI startup — said the company builds customized LLMs for each healthcare institution it serves. Typically, the aim for these LLMs is to lower costs by lessening the reliance on human medical coders. This often reduces errors in billing and speeds up repayment cycles.

"We looked at what are the biggest pain points for health systems," Walaliyadde told BI. He said that Akasa's focus is on developing LLM products for medical coding and simplifying prior authorization, a process that requires approval from a health-plan provider before a patient can receive a treatment. "Those are the ones where you could really move the needle," Walaliyadde said.

AI for health screenings

George Tomeski, the founder of Helfie AI, is in the middle of pitching investors to raise as much as $200 million in a new round of funding that he hopes to close by the first half of 2025.

Tomeski said the funding would help Helfie scale as it exits beta testing for the company's app. The app, also called Helfie, uses a smartphone camera to do medical "checks" that screen for illnesses including COVID-19, tuberculosis, and certain skin conditions.

"We're targeting all the health conditions that lead to avoidable mortality," Tomeski said, adding that the app focuses on respiratory and cardiovascular conditions. The intention is for these checks —which can cost as low as $0.20 a person per screen — to serve as a form of preventive care and as an incentive to go see a doctor in person.

While some funding is going toward sales and marketing, talent acquisition, and ensuring adherence to regulations around privacy and healthcare data, a large chunk is still being allocated to product development as AI tech advances quickly.

Dr. Brigham Hyde, a cofounder and the CEO of Atropos Health, said his latest funding announcement, in May, was timed to coincide with the geared-up launch of ChatRWD, an AI copilot that can answer doctors' questions and quickly churn out published studies based on healthcare data. Hyde said he's keen to bring in big partners this time, including the pharmaceutical giant Merck and the medical-supplies and equipment maker McKesson.

But Hyde also had to show some restraint. He said that when Atropos Health moved forward with its Series B rounds, dozens of venture capitalists expressed interest in leading the round. The company was offered up to $100 million but took only one-third of that amount.

"I don't always think that's a good idea," Hyde told BI. "As a founder, you want to raise the right amount of money for your business and for the stage you're at."

It may be tempting to take more, as many healthcare AI startups — a vast majority still in the seed and early-stage funding rounds — are racing to outmaneuver rivals. Even if the technology is right, it has to get past regulatory approvals and persuade cautious hospitals and health systems to open up their wallets.

"You can build as much product as you want, but you can never build a market," Soni of Suki said. "It shows up, or it doesn't show up."