Close

New Research from CDW on Workplace Friction

Learn how IT leaders are working to build a frictionless enterprise.

Jun 30 2026
Artificial Intelligence

Q&A: Healthcare Innovation Relies on a Strong Data Foundation

Mayo Clinic Platform COO Maneesh Goyal discusses the health system’s longstanding data-driven culture to support change in the industry.

Mayo Clinic and Microsoft recently announced a collaboration to develop and launch a frontier artificial intelligence model for healthcare, grounded in the health system’s data and insights.

The Rochester, Minn.-based healthcare organization has long established its data governance and strategy, which has been foundational to its innovative approaches. That’s why the conversations around AI for the organization have always focused on addressing problems to solve rather than keeping up with the latest technologies.

Maneesh Goyal, COO of Mayo Clinic Platform, the health system’s digital innovation offshoot, spoke with HealthTech about the organization’s data transformation and how the platform is expanding to improve care delivery across a continuum.

Click the banner below to power employee productivity with AI.

 

HEALTHTECH: What are common roadblocks for healthcare organizations in setting up the right infrastructure to turn their data into insights? How did Mayo Clinic identify what worked?

GOYAL: If you look at the journey that we took, each step had a set of obstacles that allowed us to put the building blocks in place. The first was almost 15 years ago, when we adopted an electronic health records system. That allowed us to provide our patients with a consistent Mayo Clinic experience at all our destination sites and Mayo Clinic Health System locations.

Two, we recognized that the movement from maintaining data and infrastructure on-premises to the cloud was necessary if we were going to create use cases that we hadn’t even contemplated. We made the leap well before the industry did.

The next step was to get the organization to abide by certain principles when it comes to data. One was that it is an enterprise capability or asset that we are stewarding on behalf of our patients. That concept is very important, because in most organizations, data sits in silos and is guarded just within an area of the enterprise. We took the approach that we’re organizing patient data and making it available such that our entire patient population benefits.

The next step was pushing the de-identified data into the cloud so we could do a lot more with it.

After that, we said, “Let’s make sure that our consent model is built for maximal value extraction,” and for us, value extraction means our patients getting better care driven off that data instead of it just sitting there and collecting dust. But we can only do so much. There’s a lot of change happening across the globe spurred by innovators, and we needed to figure out a mechanism to engage them. So, we created this model called “Data Behind Glass” and invited the world's innovators under a safety, privacy-protected construct to give access. We made sure that every person engaging abides by certain rules, and they’re focused on delivering value to our patients, meaning novel solutions that are focused on quality improvement, earlier disease identification or new cures. Where we continue the journey is, how do we make that an industry calling and invite others into the collaboration?

DISCOVER: Is your data governance actually AI-ready?

HEALTHTECH: How did Mayo Clinic develop its data-driven culture? What worked in encouraging stakeholders to become more data-minded?

GOYAL: Mayo Clinic is a 160-year-old organization founded on the principles that it’s not until we have multiple specialties — what we call a union of forces — come together that we get the best outcome. With other healthcare systems, typically, a patient can get bounced from one specialist to another. Our approach is, let’s put all of that in house and let those clinicians and departments confer to make the right diagnosis. That’s an important concept. So, we break apart those silos to start with.

The second concept is that we have an integrated research and clinical practice. That is probably one of the secrets to Mayo Clinic; we’re constantly pushing the envelope on care because our researchers are moving the care models and working with our clinicians. They’re not in silos. They’re integrated into each department.

Last, and it goes to why we’ve launched a platform, is that the founding Mayo brothers made it a point to teach others, but they also took what they learned from outside and brought it back and made it a core part of what Mayo Clinic is. When you think about platform models, that’s what you do: You create a common asset, and there are other people contributing and taking from it. That is the culture of Mayo Clinic. It’s a necessary component when you think about being data-centric, because it’s more of a learning organization. Data is a means — just like AI is a means — to an end. They just provide scalability. Mayo Clinic has always had these values, but the data and technologies enable us to move at a much faster pace.

Click the banner below to sign up for HealthTech’s weekly newsletter.

 

HEALTHTECH: How is AI helping to scale the platform? How did Mayo Clinic identify and focus on the business problems to solve rather than just adopt the latest technology or models?

GOYAL: So, how did we use AI? I would break it apart into technology, operations and clinical value.

On the technology side, we used AI to help us de-identify the data first in a way that protects patients. Mayo Clinic Platform has 36 petabytes of data. That’s a massive amount of data. AI tools do that at scale. Second was to then curate that data, because large-scale data sets are unstructured. Clinicians’ notes have a lot of nuggets of knowledge that need to be organized. For example, we used AI to create standardized oncology notes on top of the clinical record to stay consistent. So, curation is a necessary step.

On the operational side, we want to automate or use agentic AI to improve efficiencies, and everybody is charged with coming up with novel uses of these technologies. Our hope is that we start to create more efficiency so that we can lower the cost of healthcare and focus on delivering care. Could we shorten the time that it takes to make an appointment? Could we make the experience of prior authorization better? Can we make the experience of getting results better? AI plays a role in supporting those administrative tasks.

The last one is the one that we’re most excited about, which is, can we change the practice of medicine? We have an opportunity to consolidate information and upskill healthcare professionals, whether they’re a subspecialist or a primary care physician or a nurse practitioner. And for us, upskilling means, can we get to an earlier diagnosis? And then, can we do that at a lower cost? And we want to deliver on the promise of precision medicine. Our approach at Mayo Clinic is to figure out how to bring our capabilities to other organizations. Through Mayo Clinic Platform, we can take our unique position in the market and bring those solutions to those organizations that don’t have the same kind of business model.

RELATED: Scaling AI without scaling costs starts with token awareness.

HEALTHTECH: What’s next for the platform?

GOYAL: We’ve been thinking about how we take our model and make it an agentic model, which is, can we release tools on top of the data? If we can introduce more automation, that becomes much more scalable. Making things more scalable means we can invite partners across the system that have fewer financial resources available.

Second is to bring in all sorts of other data sets. It’s really powerful when you can bring in waveform data next to imaging data next to molecular data, because you can release tools that help healthcare professionals find new insights.

Medicine has been practiced a certain way for many years. In the age of AI, we have an opportunity to rapidly make real changes, based on additional insight that we get from data.

gorodenkoff/Getty Images