Why 96% of Healthcare Data Goes Unused — and How AI Finally Changes That | Context 2026

96% of healthcare data goes unused. AI pilots keep stalling. And most organizations are still waiting for perfect governance before they start. This session cuts through all of it. In this Track 4 session from Context 2026, Smriti Kirubanandan, Managing Director at Innovaccer, leads a grounded conversation with Ravi Koganti, CIO at Capital Health; Mike Sutten, CIO at Innovaccer; and Kalyani Gopalan, Executive Director of Analytics at Presbyterian Healthcare Services, on what it actually takes to move healthcare AI from experimentation to autonomous operations for healthcare at scale. From data strategy to AI governance, from pilot orphans to enterprise ROI, this is one of the most practical conversations of the conference. In this video: → Where healthcare organizations actually are on their AI journey today → Why context — not data volume — is the real blocker for AI in healthcare → How to build a governance framework that scales without slowing you down → The difference between AI governance and data governance — and why confusing them is dangerous → What the early signs of a failing pilot look like before it becomes an orphan → How to define ROI for AI when the returns are not always measured in dollars Chapters: 00:00 Introduction and panel overview 01:09 Where Capital Health is on its AI journey 03:04 The inflection point — from analytics to autonomous action 05:02 Where health systems are stuck between pilots and production 08:11 Why AI implementation is not a technology problem 10:30 How much healthcare data actually gets used 13:04 Why context is the missing layer in healthcare AI 17:06 Building context intentionally for AI agents 19:44 How to standardize context across a complex health system 22:34 Data governance as the foundation for AI — not an afterthought 25:09 How governance needs to evolve as AI takes action 28:22 How Capital Health built its AI governance framework 31:07 Where organizations are most and least cautious with AI 34:11 AI governance vs data governance — a critical distinction 35:20 Audience Q&A — early signs a pilot is becoming an orphan 38:46 How to define and measure enterprise ROI for AI 41:56 If you had a blank check — one AI investment for healthcare