Artificial intelligence (AI) holds tremendous potential for healthcare payers but realizing that potential hinges on one crucial factor: data readiness. In a recent webinar hosted by Emids, ‘Modernizing Data Foundations for AI: How to Accelerate Value from Legacy to Cloud’, industry leaders from Snowflake, BHI and Emids came together to explore why many payer organizations are struggling to translate AI enthusiasm into real-world outcomes, and what they can do to change that.
Fragmented Data Environments
AI innovation is advancing at a pace that outstrips many organizations’ ability to adapt. For healthcare payers, this challenge is compounded by complex, fragmented data environments. From siloed systems and inconsistent formats to unstructured and incomplete datasets, the foundational infrastructure often isn’t equipped to support modern AI workloads.
Transitioning from pilot programs to operational AI is especially difficult. Many payers struggle to define what to prioritize and how to scale proof-of-concept efforts into production-ready solutions. Without a clear roadmap and aligned data assets, progress tends to stall before value is realized.
Data Modernization: Still an Afterthought
While investments in cloud infrastructure and digital transformation continue to rise, the data that powers these initiatives is too often left behind. Managing unstructured data and fragmented clinical records is still treated as an afterthought in many payer modernization strategies.
Without a data strategy that runs in parallel with digital transformation efforts, organizations lack the ability to generate the insights AI needs to deliver value. This not only limits immediate returns but also jeopardizes the long-term adaptability required to keep up with ongoing healthcare and technology innovation.
There’s No AI Strategy Without a Data Strategy
A key mistake many organizations make is treating AI as a standalone initiative. In reality, AI effectiveness is directly dependent on data that is governed, reliable, and accessible.
Organizations must understand where their data originates, how it’s accessed, and how it is used across departments—from finance to medical management. Integrating AI into the enterprise-wide data strategy ensures consistent use of validated datasets and supports trust, auditability, and transparency across the organization.
Lack of Business & IT Collaboration
Payers embarking on AI modernization must resist the urge to pursue novelty. The most successful implementations start with solving specific, business-relevant problems. That might mean improving governance structures or bringing business teams closer to data engineers to align on real-world use cases.
Too often, AI and data initiatives are siloed within IT, with minimal input from the business side. This lack of collaboration slows down value realization and disconnects data solutions from operational needs. Engaging business users early on ensures alignment, improves adoption, and builds momentum for broader AI integration.
Legacy Platforms Block Unstructured Data
Emerging data platforms now allow for seamless integration of structured and unstructured data—including PDFs, audio files, and clinical notes—into unified environments that support richer analytics. This level of interoperability is critical for advancing initiatives such as value-based care and personalized member engagement.
Once connected to robust data foundations, AI agents can begin to accelerate analytics, automate data cataloging and quality checks, and even streamline schema generation. These capabilities bring significant operational efficiency to the enterprise.
Low Trust in Data and Partners
Trust emerged as a central theme in the panel discussion. Internally, trust is built through transparency and data lineage. Users need to understand what data informed an insight, how it was processed, and whether the output can be trusted.
Equally important is the trust placed in partners. Many AI initiatives fail not because the models themselves are ineffective but because implementation partners lack the healthcare domain expertise needed to embed solutions into complex payer workflows. Trusted partners combine technical acumen with industry-specific insight to ensure AI delivers lasting impact, not just impressive demos.
Moving Forward: Imperfect Data, Real Impact
A common misconception is that AI requires perfect data to function. In truth, no dataset is flawless. What matters more is having visibility into data quality, identifying gaps, and building processes to improve incrementally.
Success also hinges on finding internal champions who understand both the strategic and technical sides of AI. These leaders play a vital role in translating model performance into business value, balancing innovation with governance, and bridging the gap between data science and operational execution.
Aligning for Real-World AI Impact
AI’s promise for healthcare payers is real, but it will not be realized through technology alone. Payers must intentionally align AI goals with enterprise data strategy, invest in strong governance, involve business users early, and engage trusted partners with deep domain expertise. Those who start, even if imperfectly but strategically, will be best positioned to accelerate real transformation.