Webinar Insights: Interoperability, AI Innovation, and Data Governance in Healthcare

Setting the Stage: Why Interoperability Matters More Than Ever

Victor Wilson kicked off the webinar by outlining the topic’s significance: “Healthcare and life sciences organizations are at the forefront of figuring out how to store data, make it interoperable, and begin using frontier tools like AI to unlock its value.”

The session focused on findings from a joint report produced by Hakkoda and Snowflake, surveying 185 industry stakeholders. 84.7% of healthcare leaders rated interoperability as a top priority, underscoring its growing importance in advancing care delivery, improving outcomes, and integrating AI-driven solutions.

Murali Gandhirajan elaborated on the concept of interoperability in plain terms: “At the core, interoperability is about putting the patient at the center. It’s understanding what happened to the patient historically, what their situation is currently, and how wearable and patient-generated data adds another dimension to their whole health view.”

Shahran Haider reinforced this focus, explaining how interoperability supports NYC Health + Hospitals’ mission of serving all New Yorkers without exception. His organization relies on an enterprise data platform to enable analytics and AI across complex datasets, ensuring insights are actionable and accessible across departments.

Key Themes from the Webinar

1. The Three Layers of Interoperability

Shahran introduced an insightful framework for understanding interoperability in healthcare, breaking it into three layers:

• Operational Interoperability: Managing business-oriented processes, such as benefit lookups, claim payments, and prior authorizations, through standards and protocols.
• Analytical Interoperability: Enabling research and population health analytics by sharing data effectively across organizations and systems.
• AI Interoperability: The newest frontier, focusing on model-context protocols, semantic integration, and governance frameworks to ensure intelligent, accurate interactions with patient data.

Shahran explained that the evolution to AI interoperability adds significant complexity: “Semantic layers define how we standardize terms, while context layers add additional insights comparable to a human’s evaluation of the data.” This structured approach makes healthcare data actionable across AI platforms, which ultimately enhances patient care and operational efficiency.

2. Leveraging AI and Agentic AI

Another recurring theme in the session was the integration of AI technology, and especially Agentic AI, which automates repetitive tasks, enabling humans to focus on higher-value discretionary work.

Victor highlighted adoption trends uncovered in the report: “64.5% of healthcare organizations are already experimenting with AI or planning implementation, but only 33% feel they’re actually ready to operationalize it. This readiness gap is the key tension in 2026.”

Murali emphasized that successful AI adoption must start with defining business goals: “AI is not a technology-first conversation. It’s a business-first conversation. What problem do we really want to solve, what workflow do we want to disrupt, and what value will it drive?” This pragmatic approach ensures that AI solutions align with organizational priorities and deliver measurable outcomes.

3. Practical Examples of Interoperability in Action

Shahran shared operational examples illustrating how interoperability drives improvements in care for vulnerable populations and community health:

Case 1: Special Populations Program

NYC Health + Hospitals serves homeless populations through a program integrating medical care, mental health care, and support for social risk factors like housing and transportation. Previously, member enrollment data transferred manually from payers in a 5-to-7-day process. Using a data vault modeling approach, NYC H+H standardized the process internally, reducing enrollment time to 5 minutes and automating it end-to-end. This automation streamlined care coordination, enabling timely interventions for a highly complex population.

Case 2: Health Information Exchange for Patient Experience

Shahran described collaboration with health information exchanges (HIEs) to understand patient needs at the community level. For example, focusing on Brooklyn, NYC H+H assessed care gaps for patients with chronic conditions like asthma and diabetes. With real-time data sharing, the organization ensures investments in new facilities and specialized care align with community health requirements, improving patient access and convenience.

Victor praised this work as a model for leveraging interoperability to enhance continuity of care across clinics and hospitals.

4. Addressing Governance and Security in AI Applications

As healthcare organizations adopt AI, ensuring strong governance and security remains critical, especially when integrating sensitive patient data with large language models (LLMs).

Murali outlined a layered approach to governance: “Think of it as a multi-layer cake. Layered security includes restricting data access based on roles, applying row-level and column-level masking, and embedding governance guidelines deep into workflows.” He pointed to Snowflake’s capabilities as an example platform that combines granular access control with rapid yet secure data-sharing features.

Victor echoed the importance of robust governance, stating, “The more you can operate within your own construct and realm of control, the better.”

5. Moving Forward: Building the Foundation for AI Success

The speakers offered practical advice for organizations just beginning the journey toward interoperability and AI adoption.

Shahran urged attendees to treat interoperability not as a compliance project but as a strategic enabler: “The hard work needs to happen in parallel with everyday business operations. Engage executive stakeholders, explain the benefits, and secure buy-in to keep the foundation-building process moving forward.”

Victor emphasized the importance of clear problem definitions: “Instead of asking what tools we need, start by asking what problem we’re trying to solve. Work backward to define workflows and data requirements. Sometimes, working backward unveils efficiency opportunities we hadn’t considered.”