Luminai + Cleveland Clinic: Scottsdale Institute Fireside Chat, “What an Effective AI Platform Partnership Looks Like”

Sarah Hatchett, SVP and Chief Information Officer at Cleveland Clinic joined the Scottsdale Institute for a membership exclusive fireside chat on October 28, 2026. During the facilitated session, they discussed the ingredients of a successful AI platform partnership. Here are Luminai’s takeaways from the discussion.

11. Choose an AI-native partner that deeply understands healthcare operations and knows how to deploy automation across several complex, cross-system workflows

Choose an AI partner that:

• Balances being AI-native with a deep understanding of health system regulatory and operational complexity
• Earns the right to be a platform and scale across use cases by demonstrating outcomes and deploying reusable integrations, optimizing current systems, and augmenting current capabilities in order to scale faster
• Captures and leverages institutional knowledge and recognizes that every health system runs on local rules, some written in procedures and some only in people's heads
• Deploys solutions successfully using a model where forward-deployed engineers work side by side with operations teams until the intended outcome is achieved and hard-wired

22. Design the proof of concept to inform enterprise deployment

Design the proof of concept so the organization can make a clear go/no-go decision when it ends. That means focusing on a narrow problem, setting up the proof of concept to demonstrate value quickly, and defining key success criteria.

A good first use case has three traits:

1. High impact
2. A clear before and after
3. Fast time to value

Referral and order fax intake at Cleveland Clinic has these traits. The health system receives more than 5 million faxes a year, including about 1.1 million orders and referrals. Before Luminai, staff had to read each fax, match the order to a patient, enter the right information into the EMR, and route it. This was a clear place where value could be measured quickly, and model safety could be proven with humans in the loop.

The results so far:

• 770k faxes handled automatically by Luminai each year
• Average handle time dropped from 10 minutes to under 1 minute, and a 3-day backlog was eliminated, so faxes are handled the same day
• The first use case went live in about 70 days

33. Focus on earning trust before moving to full end-to-end automation

Solutions should not go straight to end-to-end automation. Start with a process that includes efficient human review, and expand automation step by step as performance and trust are established in production. Decisions should be transparent so an operator can see why a patient’s need was routed or prioritized the way it was. Exceptions should go to a person with the context needed to resolve them, both completing the task efficiently and adapting the models to handle similar scenarios better in the future.

44. Align on use case expansion that builds on early wins and leverages capabilities that were refined during initial deployment

Use case expansion can run in two directions:

1. Horizontally: more use cases within a domain or functional area
2. Vertically: new domains or functional areas

New use case prioritization should be driven by a holistic view of value. One approach is to look for hard ROI – a line of sight to reduced operating cost, lower administrative burden, or improved revenue capture – alongside softer returns in experience, safety, and quality of care.

Each deployment should build integrations, institutional knowledge, and decision logic that carry over to the next use case, so that each new use case deploys faster than the last and value compounds.

55. Keep the focus on patients and caregivers

Healthcare's capacity is constrained by a shortage of caregivers, so helping team members become more effective and efficient means more patients can be seen. We believe one main purpose of AI is to optimize access to the right caregiver, at the right time, in the right setting so that more patients can access the health care they deserve.

Read more about the partnership here

First presented in collaboration with Scottsdale Institute on October 28, 2026.