Building an AI-Ready Data Foundation: How Abacus Insights and Rhapsody Improved Identity Resolution for 13 Million Members

As health plans accelerate investments in analytics and AI, the ability to trust and unify member data has become a strategic priority. Without a reliable identity foundation, organizations struggle to support care management, risk adjustment, clinical quality initiatives, payment integrity programs, and the growing demands of AI-driven analytics.

Abacus Insights helps health plans build the data foundations required to support enterprise analytics, clinical quality initiatives, payment integrity programs, and emerging AI use cases. For a large healthcare company, that mission required integrating and governing data from claims, provider, enrollment, pharmacy, laboratory, and other core systems while maintaining an accurate view of more than 13 million member identities.

To strengthen its identity resolution capabilities and support future growth, Abacus partnered with Rhapsody to replace its existing EMPI solution. Together, the organizations executed a highly customized migration that preserved the company’s existing matching logic while improving flexibility, scalability, and performance.

The result was a modern identity foundation capable of supporting millions of member records, achieving 99.98% matching accuracy, and enabling the clean, trusted data required for advanced analytics and AI initiatives.

For health plans, accurate member identity is foundational.

“If you don’t have accurate identity resolution, everything downstream becomes more difficult,” said John Pyhtila, PhD, Chief Operating Officer at Abacus Insights. “Whether it’s care management, HEDIS measures, risk adjustment, or analytics, it all starts with having confidence in the underlying data.”

At the time, Abacus was facing growing challenges with its existing EMPI environment. Rising costs, limited flexibility, concerns about implementation accuracy, and difficulty accommodating new data requirements created obstacles to supporting both current customer needs and future innovation.

The organization needed a solution capable of evolving alongside its business, one that could support new data sources, changing customer requirements, and future growth without introducing unnecessary complexity.

Complicating matters further, this was not a traditional migration.

Rather than redesigning matching logic or implementing a new identity strategy, Abacus’s customer requested that Abacus preserve its existing matching methodology and member identity framework. The objective was simple: execute a seamless transition to a new platform while maintaining continuity across downstream systems and workflows.

“Our customer’s team feeling no impact of a solution migration was the goal,” recalled Pyhtila.[C1] 

That requirement significantly raised the stakes. Any discrepancies in member matching could impact analytics, quality reporting, operational workflows, and member-facing processes. At the same time, the platform needed to support millions of member records, strict overnight processing requirements, and rigorous security standards.

After evaluating alternatives, Abacus selected Rhapsody EMPI to serve as its new person data management partner. While matching accuracy, scalability, performance, and security were all important evaluation criteria, the decision ultimately came down to flexibility and partnership.

“From an accuracy standpoint, most solutions are very close,” said Pyhtila. “For us, the differentiator was finding a partner that could support where we needed to go next, not just where we were today.”

One example involved laboratory data management. Abacus had identified a challenge in which highly variable laboratory data could create duplicate records and introduce unnecessary complexity into downstream systems. Previous attempts to address the issue were unsuccessful.

Working closely with the Abacus team, Rhapsody proposed a new approach that would allow the organization to better manage these records while maintaining data quality and operational efficiency.

“It wasn’t a take-it-or-leave-it approach,” said Pyhtila. “Rhapsody was willing to work with us to understand the problem and think creatively about how to solve it.”

Once the decision was made, the organizations launched a highly customized migration effort focused on preserving the company’s existing matching logic while ensuring accuracy, performance, and continuity across downstream systems.

The implementation required extensive parallel testing. For months, Abacus and Rhapsody processed the same data through both environments, comparing matching results, validating member records, and identifying any differences between the two systems.

“We would run data through both platforms, compare the outputs, identify discrepancies, resolve them, and then test again,” explained Pyhtila. “That process repeated over and over until we achieved the level of confidence we needed.”

Validation also went beyond matching: the teams verified the accuracy of the demographic, enrollment, and clinical data associated with each member record to ensure consistency across downstream applications.

The collaboration made the difference. As implementation progressed, both teams worked together to accelerate testing, resolve issues, and maintain momentum toward an aggressive deployment timeline. Rhapsody also partnered with Abacus to address security requirements, implementing enhancements that aligned with the organization’s rigorous standards.

For Abacus and its customer, success wasn’t measured by implementing a new platform. It was measured by how seamlessly the transition occurred. After months of parallel testing and validation, Abacus and Rhapsody achieved approximately 99.98% matching accuracy while maintaining the company’s existing matching methodology and business rules.

The result exceeded expectations.

“We were able to explain virtually every difference we found,” said Pyhtila. “And in many cases, it was actually the Rhapsody platform returning the more accurate result.”

Just as importantly, Abacus delivered exactly what its customer asked for: a seamless transition with minimal disruption to the business.e, deliver consistent performance, and provide the resilience required for a transformation of this magnitude,” said Fergus Wills, Senior Solutions Architect, encompass. “Rhapsody gave us the confidence that we could build and scale our integration environment in a controlled and reliable way.” 

Today, the platform supports identity resolution for approximately 13-million-member records while meeting demanding operational and performance requirements.

Abacus processes large daily data feeds from multiple sources, all of which must be matched, validated, and returned within strict overnight service-level agreements. The platform also supports significantly larger monthly processing cycles without disrupting downstream workflows.

Most transactions are completed in less than 100 milliseconds, helping ensure timely delivery of data to operational and analytics environments.

“We have to ingest the data, process it, send it to our identity management partner, get it back, and then complete additional processing before delivering it to our customers,” explained Pyhtila. “It all has to happen overnight, and it has to happen reliably.”

The encompass programme is already delivering measurable operational and clinical While the project began as a platform migration, it quickly became a foundation for future innovation.

The flexibility of the Rhapsody platform has enabled Abacus to support new customer requirements and expand its identity management capabilities beyond the original implementation. The organization is already leveraging the platform to support the company’s continued growth and is exploring additional opportunities involving laboratory and clinical data management.

AsFor Abacus, one of the most important outcomes wasn’t reflected in a metric. It was the ability to move beyond a traditional vendor relationship.

“I don’t want to work with vendors,” said Pyhtila. “I want to work with partners.”

That philosophy shaped the project from the beginning and continues to guide the relationship today.

With Epic now live across all Trusts, HSCNI is focused on expanding and optimizing its digital cAnalytics platforms, machine learning models, and AI applications all depend on trusted, connected data. If information is fragmented, duplicated, or linked to the wrong individual, even the most sophisticated technologies produce incomplete or inaccurate results. That’s why Abacus views identity resolution not just as a technology function, but as a critical component of an AI-ready data foundation.

“There’s tremendous value locked within healthcare data,” said Jon Kaplan, VP of Strategic Accounts at Abacus Insights. “Organizations first need a way to organize it, govern it, and make it usable before they can fully realize the promise of analytics and AI.”

By partnering with Rhapsody, Abacus modernized its identity management environment without disrupting its customer or that customer’s 13 million members and built a scalable foundation for the next generation of data-driven healthcare innovation.