Skip to main content

Where to Build GenAI Solutions in Healthcare

We are at a crossroads when it comes to healthcare in the U.S. The cost of delivering care in this country has escalated to a degree that no other country has seen. At the same time, life expectancy in the U.S. has fallen behind that of many other developed nations. Something has to change with respect to delivering better health outcomes at a lower cost.

That said, there is some potentially good news on the horizon. Generative AI is enabling the creation of new solutions that have the potential to lower the cost of healthcare and improve health outcomes. Founders are responding by developing offerings that are already beginning to make an impact in parts of the healthcare industry. And importantly, per a recent study from Bain & Co., technology buyers in healthcare are responding by developing AI strategies and starting to purchase GenAI-powered offerings.

Here at Emergence, we’ve always kept a close eye on the intersection of healthcare and B2B software. That’s how we identified the promise of Doximity and Veeva early in their journeys. We are now seeing a number of exciting companies created that are using GenAI to tackle some of the biggest B2B problems in healthcare. It is our belief that there will be generational companies created that leverage this technology to better the healthcare industry. 

To keep pace, we’ve tapped our network of healthcare industry experts who are making tech decisions – and who will be most impacted by them – across providers, payers, and life science. These include leaders at institutions like Mayo Clinic, Blue Cross Blue Shield, Benchling, startups like League and Assort Health, and many more. They’ve shared where they’re seeing the best opportunities emerge for GenAI in healthcare and we’re excited to share their insights with you.

Below is a framework for how to think about which areas to target when building GenAI-powered applications in healthcare, rooted in our research and designed for founders deciding where they want to build.

Here’s what we heard

In our 25 expert conversations, three factors repeatedly came up that are likely to guide decision-making when it comes to GenAI in healthcare:

  • Degree of AI unlock: In healthcare, technology buyers are only interested in tech that can solve practical problems well. Decision makers want GenAI-powered offerings that deliver an order of magnitude improvement over existing solutions.
  • Urgency of customer pain: There’s no time or budget for vitamins. AI tools have to be genuine painkillers. Buyers in healthcare feel the most pain in areas where costs are very high and in places where quality needs major improvement.  
  • Perceived risk: In healthcare, most change is perceived as being risky. There is little tolerance for mistakes or inaccuracies. If there’s one thing that has the potential to slow GenAI adoption, it’s the perception of risk.

These factors each exist in degrees – nothing is binary. For now, the ripest opportunities are the ones with high AI unlock addressing areas of meaningful customer pain. It is also important to consider how buyers view the risk of an offering. However, in areas where AI unlock is significant and pain is high, we believe that buyers in healthcare are willing to experiment with offerings that may have perceived risk – provided that the companies selling them acknowledge said risk and are working on ways to mitigate it as much as possible.

Our framework

With all this in mind, we present the following framework for assessing GenAI-powered opportunities in healthcare:

Areas of greatest opportunity

As mentioned earlier, the degree to which GenAI provides a significantly better solution than previously existed – and the pain that such a solution addresses – are the two most important factors that matter to healthcare technology buyers. Below, we lay out the opportunities for GenAI in healthcare against these two factors. This gives a sense of the best near-term opportunities for founders.

A note on risk: For now, we decided to treat risk as a guide post – a lens to layer on top of the ideas you might plot on the graph above. We don’t want to dissuade anyone from working on high-risk problems, because they’re likely to yield some of the highest impact and most lucrative solutions GenAI has to offer in healthcare. That said, building solutions for the high-risk areas in our framework may require more careful planning both in terms of product, as well as in your go-to-market efforts, in order to drive adoption.

Below is a bit more detail on the areas of greatest GenAI opportunity in the provider, payer, and life science segments.


It’s clear that healthcare providers are overburdened. Doctors, nurse practitioners, physicians’ assistants, and nurses are seeing patients all day for small slivers of time. The administrative burden associated with remaining in compliance and managing billing creates mountains of back-office work. The result? Lower quality care and higher costs. Further, physicians are burning out at an increasing rate

Experts we spoke with from health systems identified the following areas as those where GenAI has the greatest ability to impact the cost and/or quality of care delivery:

  • Note-taking: Some physicians benefit from having “scribes” alongside them in patient visits – assistants who take diligent notes. Scribes ease the documentation burden that providers have to shoulder but are expensive. GenAI makes it possible to deliver human-quality notes with technology for the first time, thereby making it possible to deliver quality note-taking solutions in a cost-effective way.
  • Prior authorization: Prior authorization is the approval process that health providers go through to ensure patients qualify for coverage. It’s as thorny as it sounds and requires even more documentation (and about 5 hours of physician time every week). GenAI has the ability to automate much of the prior authorization process for providers.
  • Billing and claim management: Medical billing is complex and extremely time-consuming. Doing this right requires evaluating clinical notes from visits, applying proper billing codes, and submitting claims in the right format for each insurer. GenAI-powered billing solutions can ensure that medical billing is handled quickly and accurately.


It is incredibly complex to run a health insurance company that tracks and processes millions of claims each month. GenAI is a powerful way for payers to distinguish themselves in a crowded market and drive operational efficiency. Experts we spoke with on the payers’ side identified the following areas as those where GenAI has the greatest ability to impact their businesses:

  • Benefits education: Educating patients about their benefits is a huge cost center for insurance companies. About 80% of the calls insurers receive fall into this bucket. The custom and personal nature of these inquiries makes it hard to scale non-human solutions, but with GenAI, there’s an opportunity to break down complex questions and match people with high-quality answers. 
  • Prior authorization: This is just as much an issue for providers as it is for payers. We see a major opportunity for founders to develop sophisticated prior authorization solutions that streamline this process for health insurers.

Life Science 

Life science is a hits-driven business. It’s all about finding the next life-changing drug, getting through the regulatory process, and then bringing it to doctors and patients. The average drug takes 12 years and $2.3 billion to bring to market (up 15% since just last year). Further, the life science industry is under consistent pressure to lower prices even as their own costs rise. 

Experts we spoke with from life science companies identified the following areas as those where GenAI has the greatest ability to impact their businesses:

  • Clinical trial recruitment: Testing a drug demands a sample group with very specific attributes, symptoms, and experiences. In order to find them, researchers need to delve into unstructured data collected from thousands of sources. This is the ideal task for GenAI, which could identify patterns in population data, speed read through records and point pharmaceutical companies directly to the ideal patients for trials.
  • Clinical trial optimization: 90% of clinical trials fail for reasons ranging from poor study design to unrealistic deadlines. GenAI could be used to simulate and debug trials before they start and monitor compliance once they get going. 
  • Clinical trial reporting: If GenAI is great at anything, it’s turning a bunch of disparate data into prose. This is exactly what’s needed to report on the results of clinical trials. Speed is crucial at this point in the process because the clock on drug patents starts immediately after they’re identified. 20-year patents are commonly halved by testing and reporting, and every day that passes is millions of dollars lost. If GenAI can generate FDA-ready reports faster, it could have an outsized impact on bottom lines.
  • Marketing campaigns: Extensive medical, legal, and regulatory (MLR) reviews are a prerequisite for promoting any pharmaceutical product. Compliance is critical. GenAI could potentially automate a lot of the MLR process and help companies create compliant materials and messages more quickly – again saving time and nontrivial profits.

Some watch-outs and takeaways

While every expert we spoke to was wildly enthusiastic about the promise of GenAI in healthcare, we’d be remiss if we didn’t mention some of the caveats we heard in our conversations: 

  • Risk of commoditization: Given that many of the opportunities we’ve identified can be addressed with off-the-shelf or open source GenAI technology, it’s certain there will be a number of competing solutions solving these problems. There is risk that solutions in these areas may become quickly commoditized.
  • Outsourced competition: Many of the jobs to be done listed above are currently addressed by customers using outsourced, human-powered services. While these services aren’t perfect – and they are often expensive – they do work. In order to win, AI solutions will need to be not just a little, but a lot cheaper (as well as potentially more effective) than outsourced options.
  • EHRs are in the race: Heavy hitters in electronic health records (EHRs) like Epic and Oracle Cerner are not sleeping on GenAI. They’re already beginning to invest in GenAI-powered features, and with their broad adoption (the two mentioned own 63% of the market combined), have meaningful mindshare with technology buyers. Beating them will require creating superior products.

    None of these potential challenges should deter founders from building GenAI-enabled solutions for the healthcare industry. The scale of the problems to solve in healthcare is massive. The customer pain is significant. There’s unprecedented momentum and buy-in among prospects. And further, the tech itself is capable of new things every day. Those who move fast to tackle overt customer pain with tech that has demonstrable ROI can build generational companies. 

    We hope the above frameworks are useful, whether you’re determining what to build or considering how to bring AI into your existing product set. If you’re a founder building at the intersection of GenAI and healthcare, we’d love to keep this conversation going. You can reach us at and