16 Sep PODCAST: The Need for AI Ready Data in Healthcare

In our latest episode of Advancing Healthcare Through Data, Mark Coetzer, VP of Business Development at IMAT Solutions, joins us to discuss one of the most critical issues shaping the future of AI in healthcare, which is the urgent need for AI ready data.

AI is often seen as the future of healthcare, but its real potential can only be realized with clean, complete, and integrated data. As Mark explains, without high quality input, even the most advanced algorithms will fail to deliver meaningful results. This is a challenge many health systems face today because of fragmented sources, inconsistent formats, and outdated systems.

In this episode, we explore how healthcare organizations can move from “garbage in, garbage out” to building a strong data foundation that supports both clinical and financial transformation.

• Why AI ready data is essential for modern healthcare. (1:00)
• The real-world impact of fragmented, inconsistent, and incomplete data. (1:44)
• The multi-step process for achieving AI ready data through integration, normalization, enrichment, validation, and governance (3:50)
• How IMAT Intelligence helps automate and scale these processes to deliver real-time, usable data. (4:55)
• A practical example of using AI ready data to better manage diabetic patients. (5:34)
• Why the new CMS HCC V28 risk model makes accurate and normalized data more important than ever. (6:21)

In addition, as Mark highlighted in the interview, he has written about this topic for publications such as Health IT Answers, Healthcare Business Today, and HIT Consultant.

Ready to turn your healthcare data into a true strategic asset? Contact us today to learn how IMAT Intelligence can help.

Don’t have time to listen to the podcast? You can read the full transcription of the interview below:

Welcome to the Advancing Healthcare Through Data podcast series where we explore the future of healthcare and the innovations that are shaping it. I’m Matt Langan, your host, and today we’re diving into a topic that’s at the heart of the AI revolution in healthcare: data. Specifically, the critical need for AI-ready data.

And to help us unpack this, we are joined by Mark Coetzer, VP of Business Development at IMAT Solutions, a leader in healthcare data integration, who will break down how to make healthcare data AI ready.

Mark: Thanks for having me, Matt. It’s a pleasure to be here.

Matt: So, Mark, we hear a lot about AI’s potential to transform healthcare – from predicting diseases to personalizing treatments. But what’s this concept of “AI-ready data,” and why is it so important?

Mark: That’s a great question, Matt. Essentially, AI-ready data is clean, complete, and comprehensive data that’s ready for machine learning algorithms to process. Think of it as the fuel for the AI engine. Without high-quality fuel, even the most advanced engine will sputter and fail. In healthcare, this is often the case. We have a “garbage-in, garbage-out” problem.

Matt: “Garbage-in, garbage-out” – that sounds serious. Can you elaborate on what that means in a clinical context?

Mark: Absolutely. Healthcare data is notoriously messy. It comes from various sources – electronic health records (EHRs), labs, imaging systems, wearables – and each source has its own format. This leads to data that is fragmented, inconsistent, and often incomplete. Imagine trying to train an AI model to predict patient risk when the data it’s learning from is full of errors and gaps. The predictions will be unreliable, and in a worst-case scenario, harmful.

Matt: So, it’s not the algorithm that’s the problem, but the data itself. I know you and IMAT Solutions have been vocal about this. I believe I read a couple of articles you authored on this very topic, one in Healthcare Business Today and another in HIT Consultant, where you really emphasized that data infrastructure is the key.

Mark: That’s right, Matt. We believe it’s crucial to get the word out there. The industry has been so focused on the exciting potential of AI algorithms that the foundational data problem has often been overlooked. In those articles, I really wanted to highlight that a solid data foundation is the difference between a one-off AI proof-of-concept and a truly scalable, sustainable transformation in healthcare.

Matt: So, what is the solution? How do we build that solid foundation and get from this “garbage” data to “AI-ready” data?

Mark: It’s a multi-step process. First, you need to integrate data from all those disparate sources. Then, you must normalize it, which means standardizing the data so it’s consistent across the board. The next step is enrichment, where we add clinical context to create a complete patient story. Then comes validation, where clinical data specialists ensure accuracy. And finally, you need strong governance to protect patient privacy and comply with regulations like HIPAA.

Matt: It sounds like a complex undertaking. Is this where IMAT Solutions comes in?

Mark: It is. We’ve just launched our new IMAT Intelligence product line, which is designed to address these very challenges. It’s a comprehensive data management platform that automates the process of cleaning, normalizing, and enriching clinical data, making it AI-ready. We’re essentially building the solid data foundation that healthcare organizations need to succeed with AI.

Matt: Can you give us a real-world example of how AI-ready data can make a difference?

Mark: Of course. Let’s take a health plan that wants to proactively manage diabetic patients. With AI-ready data, they can use predictive analytics to identify patients at high risk of developing complications.1 They can then intervene early with personalized care plans, preventing serious health issues and reducing costs. This is the kind of proactive, preventative care that AI promises.

Matt: That’s a great clinical example, Mark. But what about the financial and operational side? We’re hearing a lot about big changes coming from CMS for reimbursement. How does AI-ready data play a role there?

Mark: You’re likely referring to the new Hierarchical Condition Categories, or HCC, model V28, which goes into full effect in 2026. It’s a significant change in how CMS calculates risk scores for Medicare Advantage patients, and it’s a perfect example of why AI-ready data is no longer a luxury, but a necessity.

Matt: And that risk score directly impacts reimbursement, right?

Mark: Exactly. Accurate ICD-10 diagnosis coding is pivotal to managing that risk adjustment.2 Under V28, thousands of ICD-10 codes are being removed from the model, while the clinical definitions for others are becoming more stringent.3 Organizations that can’t accurately capture and code the complete health status of a patient will face significant reimbursement challenges.

Matt: And this is where AI-ready data becomes a financial tool?

Mark: Precisely. With a clean, comprehensive, and normalized dataset, you can leverage AI and Natural Language Processing, or NLP, to scan through all patient data—including unstructured clinical notes—to identify documented conditions that may have been missed or improperly coded. This helps organizations close coding gaps, ensure compliance with the new V28 model, and ultimately protect their revenue. It turns data from a simple record into a strategic asset for financial health.

Matt: A future of smarter care and more stable finances—that’s a powerful combination. Mark, thank you so much for shedding light on this critical topic.

Mark: My pleasure, Matt.

This concludes this episode of the Advancing Healthcare Through Data podcast, where Mark Coetzer, VP of Business Development at IMAT Solutions, a leader in healthcare data integration, broke down how to make healthcare data AI ready.