Enterprise AI for Health Plans: A Fireside Chat with HealthEdge® CTO Rob Duffy

AI is changing business operations across industries. See how HealthEdge® is integrating AI to automate workflows & improve business outcomes.

Artificial intelligence (AI) is reshaping how industries operate. It’s not just a transformational opportunity for health plans – it’s an urgent one. Rising costs, labor shortages, increased compliance requirements, and administrative complexity are pushing plans to rethink how work gets done, and AI offers a way forward. We sat down recently with Rob Duffy, Chief Technology Officer at HealthEdge®, to explore what becoming an AI-native enterprise means for health plans and the partners who support them.

Rob shares his vision for reimagining the structure of work with AI, his approach to responsible AI adoption, and why the most significant breakthroughs in healthcare will come not from front-end tools but from transforming the everyday processes that quietly power the system.

Generative AI is currently a catalyst for innovation in many industries. What do you see as its most transformative applications for health plans over the next few years?

Rob Duffy: People often jump straight to front-end applications or member-facing tools when talking about AI. And those are important. But the most transformative potential lies in how we refactor work itself.

Across the healthcare industry, many processes still rely on long, manual sequences— read this, look that up, log into three systems, extract five data points, and re-enter them somewhere else. A single task can require 20 separate steps that take up time and create bottlenecks. Agentic AI  refers to AI systems and models that can act proactively and autonomously to achieve goals. Agentic AI can take a meaningful subset of those steps off your plate.

Imagine not having to check multiple systems to verify one claim or read 30 pages of documentation to find the two sentences that matter. AI can summarize, extract, auto-complete and present the information you need when you need it. The result? If you’re a care manager, you have more time for members. If you’re in operations, you can resolve backlogs faster and more accurately.

Ultimately, this kind of transformation frees up human capacity for higher-order work and drives better digital experiences. But step one is about reducing friction. We need to start by eliminating 40 to 50 percent of the redundant steps people are still performing across the enterprise. From there, the real innovation can begin.

HealthEdge is on a mission to become an AI-native enterprise. Can you give us a high-level overview of what this transformation entails?

Duffy: Think about the cloud migrations we’ve all gone through over the past decade. Those efforts had structure. We created centers of excellence, developed frameworks, categorized workloads using frameworks like the 6 Rs: rehost, refactor, rearchitect, rebuild, retire, and replace. And we used those for every workload to say, “How will we treat this workload in the cloud?”

That kind of framework doesn’t exist yet for AI. What most organizations are doing now is experimenting. They’re handing people tools like ChatGPT or Copilot and saying, “Try it out.” That can be useful in the short term, but it won’t drive systemic transformation.

At HealthEdge, we’re flipping that approach. We treat AI transformation like a major migration effort. We’ve created an Agentic Center of Excellence and developed our own model for identifying, mapping, and migrating work. We then move work through our version of a factory, just as we would with infrastructure. This time, however, the work we’re mapping encompasses human activities rather than technical systems – tasks, decisions, interactions, clicks, and manual reviews are all in the hands of people.

We ask: Can this be eliminated? Can this be automated? Can we augment the human in the loop instead of replacing them outright? Once we evaluate it, we push it into what we’re calling the “agentic factory,” a structured, repeatable way to move work from human execution to AI systems.

But agentic AI isn’t about replacing people – it’s about partnering with them and augmenting how they work. We need to give people a model that helps them see what AI is doing, why it’s helping, and how they’re still in control.

That’s why HealthEdge is being intentional. We’re not just putting technology out there, but building governance, guidance, and adoption models that help teams know how to use it, when it’s appropriate, and how to build trust.

As our transformation progresses, HealthEdge is focused not only on internal enablement but also on establishing a blueprint that others in healthcare can follow. By leading with process discipline and repeatable models, the company is making it easier for health plans to adapt and scale responsibly.

The industry doesn’t need more experimentation. It needs scalable progress. We are focused on achieving real business outcomes, reducing friction in high-effort workflows, and improving the delivery of care. That means embedding AI directly into our core digital healthcare platforms, not layering it on top. When we do that, we can streamline operations, improve the member experience, and drive real value for our customers.

That’s the kind of impact we aim for: not just a smarter tool, but a smarter system. That’s what HealthEdge is building.

Can you tell us about your agentic AI vision for HealthEdge?

Duffy: My vision is simple: stop doing proofs of concept and start doing real work. We’ve reached a point where experimentation alone isn’t enough. If the only AI in your life is a robot vacuum that sweeps the kitchen floor, you’re not tapping into the full potential of agentic systems.

Our goal is to identify the repetitive, low-value tasks no one wants to do and automate them with intelligence, not just with scripts or bots, but with systems that can reason, respond and learn.

For us, this isn’t about one use case or one solution. It’s about building the infrastructure and processes to apply AI across the board, with governance and accountability built in from the start. That’s how we get to scale, and that’s how we deliver real results for health plans, providers, and members.

Healthcare is a notoriously complex and risk-averse industry. How is HealthEdge balancing innovation and accountability to ensure trust, security and fairness in its AI initiatives?

Duffy: Innovation only works when it’s trusted. To that end, we’re taking a measured and structured approach. First, we’ve developed an internal AI adoption policy and enterprise risk governance model. It draws on frameworks like the NIST AI Risk Management Framework and is guided by principles of fairness, transparency, safety and regulatory compliance.

Second, every candidate technology goes through a structured evaluation process. We don’t deploy a tool just because it’s available. We assess its performance, examine its risk profile, and determine whether it aligns with our internal standards and our customers’ values.

Third, we have a governance council to oversee this work and provide input on direction and oversight. Internally, we want to become excellent at managing innovation responsibly so that externally, we can lead with confidence and transparency.

Our goal is not just to innovate, but to do so in a way that earns and maintains trust. If we get that balance right, the rewards for our customers will be enormous.

For other organizations in healthcare looking to adopt AI, what strategic advice would you offer to help them achieve impactful outcomes while avoiding common pitfalls?

Duffy: Treat AI adoption like a transformation project, not a side experiment. Too often, organizations pursue whiz-bang use cases because they’re flashy or popular. But those don’t always translate into meaningful impact.

I recommend starting with a work inventory. Map what your teams are doing. Look for places where tasks are repetitive, rules-based, or require high cognitive load but low judgment. Then, apply the same rigor you would to a cloud migration. Build a repeatable model. Plan it. Staff it. Execute in waves.

If you do this right, you’ll find that 15 to 20 percent of your people’s work today could be streamlined, automated or augmented. That’s not a hypothetical number. That’s an actual opportunity. It’s not about replacing people. It’s about giving them better tools and freeing them to focus on higher-value work.

What excites you the most about the future of AI in healthcare?

Duffy: There’s a quote I love from Michio Kaku: “When a technology becomes sufficiently advanced, it becomes both everywhere and nowhere.” He used electricity as the example. It’s in everything we do, but we never really notice it. We just expect it to work.

That’s what I want for AI. I want it to become so embedded in our systems, so well-integrated into the experience of delivering and receiving care, that it fades into the background. That’s when you know you’ve really made it.

Imagine a world where care managers don’t have to dig through files to understand a member’s history, claims are processed instantly with no manual review, and members receive proactive outreach because AI knows when they need support. That world is coming, and we’re building toward it now.

Ready to transform your health plan with enterprise AI?

Explore HealthEdge’s Enterprise AI strategy to start your AI transformation today. Contact us to get started.