Dr. Ewa Kleczyk shares why responsible healthcare AI depends on human oversight, organizational trust, and an honest understanding of model limits.
About Dr. Ewa Kleczyk
Dr. Ewa Kleczyk is Chief Data & Analytics Officer at Prolaio, a cardiovascular digital health company, where she leads the Data Science & Services organization spanning digital biomarkers, health insights, and data services. With more than two decades of experience in healthcare analytics and research, she has held leadership roles building expertise in real-world evidence, health economics and outcomes research, clinical trial analytics, and digital biomarkers. She serves on the Advisory Committee for Virginia Tech and is an affiliated graduate faculty at the University of Maine, and holds an executive-level role with WomenTech Network. Alongside her husband, she supports several non-profit organizations focused on education and equitable healthcare access. She is a published author, book author on leadership, and frequent speaker on AI, data strategy, and organizational leadership.
What motivated you to join the AI Builders Global Conference?
Absolutely. What drew me in is the conference's focus on builders rather than just theorists, people who are actually putting AI into production and living with the consequences of those decisions. As someone leading data science and analytics in cardiovascular digital health, I spend my days in that exact tension between what a model can technically do and what an organization is actually ready to trust and act on. I wanted to be part of a community wrestling with that honestly.
What brought you to healthcare analytics and AI?
My path into healthcare analytics and AI wasn't linear, and I've come to see that as an asset rather than something to smooth over. Over nearly two decades, I've worked across real-world evidence, health economics and outcomes research, and clinical trial analytics. What has consistently pulled me forward is the stakes involved: in health care, a model isn't a demo, it's a clinical decision support tool that clinicians and patients are relying on. That responsibility is what made me fall in love with this field, and it's shaped how I think about leadership, with an emphasis on authenticity, psychological safety, and making room for non-linear career stories like my own.
Why is this conversation important now?
Because the hardest part of AI work is almost never the algorithm, it's everything around it. A model can be technically excellent and still fail if clinicians don't trust it, if an organization hasn't built the workflows to act on its output, or if the team building it doesn't have the discipline to say clearly where the model stops being reliable. In healthcare, that discipline is not optional, it's a patient safety issue. I want to talk candidly about what it actually takes to earn organizational buy-in, keep humans meaningfully in the loop, and resist the temptation to oversell what a model can do.
Who should be part of this discussion?
I'd point to two groups in particular: technical leaders who are past the proof-of-concept stage and are now grappling with deployment, trust, and governance; and leaders in regulated, high-stakes industries, healthcare especially, who need practical, honest conversations about where AI genuinely helps and where human judgment has to remain firmly in charge. This isn't a conference for hype; it's for people doing the harder work of making AI actually usable and accountable.