Editor's Note: This is the seventh in our series on Chief AI Officers in Healthcare. Other recent CAIO profiles include Dennis Chornenky at UC Davis Health, Dr. Karandeep Singh at UC San Diego Health, Alda Mizaku at Children's National Hospital, Dr. Zafar Chaudry at Seattle Children's, Mouneer Odeh at Cedars-Sinai and Sameer Sethi at Hackensack Meridian Health.
As chief data and AI officer at Ann & Robert H. Lurie Children's Hospital of Chicago, Rajiv Kolagani is building a state-of-the-art data platform that harnesses artificial intelligence to elevate clinical quality, drive operational excellence and enhance the experience of providers, patients and families.
Kolagani's 20-year career features pivotal leadership roles at organizations including Duly Health and Care, DaVita and Centene, where he successfully scaled Medicare Advantage programs, implemented risk-based reimbursement models and deployed AI-driven population health analytics that demonstrably improved patient outcomes.
Now he's got one of the newest and fastest-growing titles in healthcare: Chief AI Officer.
"A couple of years ago, Lurie Children's Hospital was looking for someone who could come in and transform the data and analytics and AI program," says Kolagani. "It realized the need for transformation in the data space, mainly because the technology and the operating model were outdated.
"There were issues with trust with data and adoption," he continued. "So that's when they tapped my shoulder to say, 'Hey, do you want to explore this opportunity to really transform data and AI and analytics?' Over the last two-and-a-half years, there has been a significant amount of investment made to transform people, process and technology here at Lurie, from a data and analytics perspective. We've modernized our entire technology stack."
Building a team, finding value
From a people perspective, the organization has added a few individuals who focus mainly on best data engineering practices and services. Kolagani has a senior director who focuses on that area. He also has a director of AI, a senior director of product, and a senior director of data and AI literacy.
"And then on the process side, we've moved away from a pure ticket-and-service model to product and internal product and agile delivery models that have significantly increased how we engage with our customers and how they receive value in our program," he explained.
"The last six to eight months have been really focused on getting AI and the strategy for AI and the governance for AI correct," he continued. "What we realized is we needed a leader at the enterprise level who can orchestrate all of the AI within the organization, govern it centrally, but execute it more locally."
That's been the mantra because there's not one person who should be doing all the AI, Kolagani added.
"AI should be happening across the entire enterprise," he noted. "So that's when they tapped me to be the Chief AI Officer. I report to the chief digital and information officer. I've essentially been doing the AI role for about one year, but the official title came in March 2025."
Tech and data, strategy
Kolagani said he always has had a knack for setting strategy. He started in the technology space, and that's been where his growth first happened, for the first five years of his career. But after he earned his MBA from Duke, he took on roles where he became the bridge between technology, strategy, and data and analytics. That prepared him for leading strategic roles.
"When you think about AI strategy and technology strategy, they really need to dovetail nicely into business strategy," he said. "They should not exist separately from business strategy. Understanding the provider business, understanding the research business, understanding healthcare in general, really has prepared me to take on the Chief AI Officer role.
"From an AI perspective, I've been doing AI for the last 10 years or so on the provider side and on the vendor side," he added. "When I was working with Press Gainey, we were working with large data sets. We would bring in clinical quality, safety, and patient experience and employee experience data for 80% of healthcare in a very large benchmarking database."
Then Kolagani would find correlations using machine learning between all of those data sets and figure out what drives outcomes, what are the leading and the lagging indicators, and what's the actual correlation between the two.
"When I was with DaVita, this was a few years ago, we were predicting disease progression from a chronic kidney disease perspective," he said. "That was purely on chronic kidney disease data.
"Now, generative AI is newer, but the AI discipline is much broader than generative AI," he continued. "The machine learning aspect of it and what we call classical AI, I've been doing it for quite some time in the healthcare space. I've done a bunch of work in the payer space, too."
Not a 'one-person thing'
Kolagani said IT executives looking to become a Chief AI Officer need to have some technical AI chops, just to understand the basics about AI. But it's really about understanding strategy – aligning AI and technology advances to the business strategy. And bringing people together and to collaborate on long-term objectives.
"AI is a team sport, it really cannot be a one-person thing," he said. "What I would recommend to leaders who want to do more AI is really understand the capabilities and the technology, understand what it can do, understand its potential, understand the limits of this technology. You really have to figure out when you need to put the human in the loop and when you let AI run what it needs to do.
"It's an art and a science," he continued. "Once you master that, when you really need to put the human in the loop, you're going to have a ton of benefits that are going to come about. I would say technical skills but also understand what the business problems are that you're really trying to solve."
A Chief AI Officer must understand the business problems and the key challenges to be addressed and map out a strategy to go after them, he added.
What a CAIO does
Broadly speaking, Kolagani's role as Chief AI Officer is to organize AI for Lurie Children's Hospital. First, activating AI within the organization. What are all the different capabilities to activate? How do leaders advise on the right capability to use? Is it build or buy?
Then there is AI literacy. How do leaders educate on the power of AI? The capabilities being launched. The benefits the capabilities will have within the organization.
"Now, what that looks like is we have enterprise AI, which is all of the things that Microsoft and Workday and all of these platforms provide," he explained. "We have clinical AI, which is Epic and the clinical systems and the functionality they have. Then we have what we call the bespoke AI or custom AI we're developing, that enterprise and clinical AI do not provide for.
"We're having to build these custom systems ourselves," he continued. "For enterprise and clinical, we work with our enterprise partners on activating these capabilities and moving those through the AI governance model and the approval process."
Then, the team develops all of the custom AI systems.
Will the model be effective?
"So, on a typical day, it could look like we're running governance in the subcommittee one day, we're running advisory one day, which is basically intake of all of the AI capabilities and understanding which ones are going to create value," Kolagani said. "And then we run them through an entire process of security review, legal review, risk review, AI model review, to really understand, 'Is this model going to be effective?'
"And then the full ROI model for each AI capability," he continued. "And then, does this need to go through a pilot process? Is it ready for production? How do we implement these AI capabilities?"
Kolagani and team thus must cover the entire lifecycle of AI. He has consulting teams and internal teams that meet with business owners and a variety of stakeholders to make this whole process work. They run the governance, advisory, literacy and adoption processes.
One AI project Kolagani is proud of is a custom system built by him and his team that understands hospital-acquired conditions such as central line-associated bloodstream infection, surgical site infection and others.
"It's understanding how the infection happened using data," he explained. "The typical processes for an infection preventionist, they would sift through volumes and volumes of data. If we have a patient who has a length of stay of one to three months, you're talking about almost five to 10 textbooks' worth of medical records you would have to go through.
"But we found that generative LLMs work very effectively when we deploy them at reading through the entire body of knowledge," he continued. "They can read through the entire chart, and we can ask questions. So, we've built an intelligent assist agent that sits on top of the patient's clinical record. And you can just ask questions."
What's the AI looking for?
For example: When did this infection event happen? What was the microorganism you found on this particular day? What was the temperature of the patient at that particular event?
"And it pulls all of that," Kolagani said. "Some of the infection preventionists have said they would take hours to days to review some of these complicated charts, and they're able to bring that down with AI to just minutes. In less than an hour, they can do the full analysis of the chart just by asking questions.
"And this is where the power of LLM, the power of AI, really comes to life, because we found a real problem where it is an administrative burden, where their time now can be deployed to doing other things," he added. "They can round effectively, they can educate, they can be proactive with these infections."
Kolagani and his team developed all of this in less than three months. He credits AI and IT work done prior to building this system.
"We pulled all the data to the right place," he explained. "We put it all on the right platform where it has both AI and data capabilities. And guess what happened? We were able to just build these LLM modules on top of that data because we know we have a good data set, we have a trusted data set, and we're able to pull this type or build this type of capability for our end users.
"And they couldn't be more happy," he continued. "Now they just want more. Now, we've opened the floodgates. Can this product do this? Can this actually work on these other infections? We also do chart review for a bunch of different complicated conditions. I'm very proud of my team and very proud of the results and the outcome."
Click here to watch a short video of bonus content, where Kolagani describes where IT execs should start with generative AI – and where they should be cautious.
Follow Bill's HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.
WATCH NOW: Generative AI advice from a top Chief AI Officer
