Building a healthcare AI masterpiece, brick by brick

Imagine building a complex LEGO® city. Each piece, no matter how small,
plays a crucial role in creating a cohesive and functional masterpiece.

The same principle applies to healthcare data and artificial intelligence (AI). Too often, we see healthcare organizations overlook critical connections or underinvest in the business architecture and transformation necessary for success. To address these gaps, we identified six essential building blocks for a robust healthcare data and AI ecosystem.

Let’s dive into each block and see how they fit together to facilitate
connected, intelligent, and adaptive healthcare organizations.

1. The blueprint: Operating model design

Just like a LEGO instruction manual, the operating model provides the blueprint for an organization’s data and analytics culture, talent, structure, governance, and ways of working. At Slalom, we emphasize a customer-centric approach to operating model design, starting with alignment with the organization’s overall strategy and priorities. This requires early and consistent engagement with clinical, operational, and business leaders to maintain focus on outcomes.

2. The foundation: Enterprise data platforms

The data platform is the foundation of the healthcare data and AI ecosystem and must support the entire structure. Modern healthcare data platforms collect, store, manage, and share interoperable data to enable real-time and longitudinal data workflows. Unfortunately, data platforms can be a significant investment for healthcare organizations. A common mistake we see is organizational leaders trying to reduce costs by investing in “lift and shift” migrations. Imagine you’re building a LEGO city:

• Lift and shift is like picking up your entire LEGO house and placing it on a new table. This approach appears technically feasible on the surface and has lower up-front investments, but it usually results in significant cost overruns and unmet expectations.
• Replatforming is like moving the house but swapping out the baseplate for one with better features—some improvements, but the structure mostly stays intact. This can be a good first step toward modernizing your data platform.
• Re-architecting, on the other hand, is like taking the house apart and rebuilding it with new bricks and a design to better fit the new environment—it takes more time and investment, but you end up with something stronger and more adaptable.

Regardless of your approach, some guiding principles to consider are maintaining a relentless focus on clinical and operational use cases, designing in security and governance controls, and automating everything—or at least as much as possible.

3. The building blocks: Healthcare data models

Before you build your healthcare data city, you need to organize your LEGO bricks—by color, size, and function. Data models define how your bricks are structured and related. Each piece must fit perfectly to create your LEGO city. Some things to consider when sorting your bricks:

• Design your master plan (conceptual design): Draw a map of your healthcare data city, including your core entities (e.g., patients, encounters, claims, labs) and the relationships among them.
• Sort and label your bricks (logical model): Specify what type of bricks you need and how they’ll click together. This includes identifying data standards (e.g., ICD-10, SNOMED, OMOP) for compatibility and creating an entity-relationship diagram (ERD) to provide the blueprint for construction.
• Model for reuse and scalability: Smart builders don’t glue their LEGO
  sets together. Instead, they build modules—self-contained sections that can be reused or rearranged. This requires reusable components, auditability, and metadata for governance, ownership, and usage rights.

Healthcare data models enable you to build a LEGO ecosystem where every set clicks with every other, every piece has a story, and expansion is seamless.

4. The command center: Reporting & insights

You’ve built the framework, now it’s time to activate the command center. This is where your healthcare data builders, residents, and visitors access the data necessary to operate your city. Here’s how you get your command center fully operational:

• Standardized data extracts: Regulatory reporting can be a huge burden on your builders and residents. Predefined data extracts that comply with your LEGO building requirements will help you generate and submit reports, maintaining compliance and avoiding penalties.
• Self-service tools: Running a modern LEGO city requires timely and accurate information, along with complex analytics for future planning. Dashboards provide LEGO city leaders with real-time information about operations, finances, and safety to make day-to-day decisions, while sandboxes give city workers safe, isolated environments where they can explore and test data without destabilizing city infrastructure.
• Real-time data exchange: Not all your LEGO citizens speak the same language, and your workforce must collaborate with neighbor communities and other organizations to run smoothly. Fast Healthcare Interoperability Resources (FHIR) APIs enable real-time data exchange between different healthcare systems to ensure up-to-date citizen information is available across various platforms.

5. Innovation districts: Generative & agentic AI

Imagine that within your LEGO city, you are building new urban innovation hubs where your employees and citizens can generate insights, create city designs, and even draft blueprints for future buildings and transportation networks. In these districts, your LEGO team members work in hybrid pods with autonomous agents to streamline operations and enhance decision-making.

Your mayor just announced a new Healthcare Innovation District, where the residents can ideate, pilot, and scale use cases for clinical and business operations. Some of the use cases in the hub will include:

• Generating doctors’ notes and personalized care summaries.
• Drafting prior authorization requests or appeals letters.
• Serving as a digital assistant to clinicians and administrators.

The district will operate under an enhanced city code based on guidance from the Coalition for Health AI (CHAI) for responsible AI practices. Targeted investment and cross-functional leadership are critical in these zones, so you appoint a chief intelligence architect to oversee the design and evolution of the district.

6. The master builder: Healthcare AI orchestration

Finally, every LEGO masterpiece needs a master builder—someone who sees the big picture and ensures all the pieces come together seamlessly. AI orchestration transforms a collection of data, tools, and agents into a truly intelligent system by:

• Connecting the infrastructure: Integrate data pipelines, APIs, and services across systems.
• Coordinating complex builds: Manage workflows that span multiple AI models and decision points.
• Monitoring the city’s pulse: Check performance, drift, and outcomes to drive continuous improvement.
• Bridging city departments: Enable cross-functional collaboration between data and analytics teams, clinicians, and business leaders.

In a world where your zip code can have as much impact on healthcare outcomes as care providers, AI orchestration has the potential to help connect health and human services so that even the most vulnerable LEGO citizens can enjoy wellness and vitality.

Conclusion

Just like building a LEGO masterpiece, creating a data-driven healthcare system requires a clear vision, careful planning, and the right pieces. At Slalom, we are committed to partnering with healthcare organizations to build and implement these foundational elements, driving innovation and outcomes.

Let’s build something extraordinary—together.