Scaling Population Health with AI While Preserving the Human Touch

Summary

---

Episode Overview

Title: Scaling Population Health with AI While Preserving the Human Touch
Date: May 19, 2026
Host: Grace Lynn Keller (Becker’s Healthcare)
Guest: Betsy Williamson, R.N.B.S., MHA – Vice President of Quality Performance and Population Health at Medical Mutual of Ohio

In this episode, Grace Lynn Keller hosts Betsy Williamson to discuss how Medical Mutual of Ohio is leveraging AI—specifically, an AI voice agent named “Stephanie”—to scale population health initiatives while maintaining their commitment to personal, human-centered care. Williamson details their experience, addresses resistance to AI both internally and externally, and shares real outcomes and lessons learned throughout implementation.

---

Key Discussion Points & Insights

1. Medical Mutual’s Unique “Hometown” Identity (01:12–01:50)

• Medical Mutual is a mutual company, owned by its members rather than Wall Street. This enables a closer connection to the communities and a focus on serving the health needs of fellow residents.
• Quote:
  “We live in the same neighborhoods as our members do, we use the same providers, we are shopping in the same stores. So we really understand our members' needs and what is happening within the community which allows us to really care for our members in a better way.”
  — Betsy Williamson, 01:24

2. Balancing AI Innovation with Human Touch (01:50–02:56)

• Despite Medical Mutual’s brand promise (“you call us, you get a human”), they adopted “Stephanie,” an AI voice agent, not to replace, but to augment their care teams.
• Key Idea: AI is used to expand reach and efficiency, with human support always available for escalation.
• Transparency is a priority: members are informed in advance when they’ll be contacted by Stephanie and can always opt to speak to a live person.
• Quote:
  “Stephanie does not replace humans... she actually augments it. So, you know, we're able to expand more and reach more people and help them through their journey.”
  — Betsy Williamson, 02:34

3. Change Management: Overcoming Internal Resistance to AI (02:56–04:55)

• Williamson emphasizes that adopting AI is inevitable (“the train’s already left the station”) and that staff buy-in relies on clear communication, transparency, and comprehensive change management.
• Staff are reassured of their critical role, given clear escalation processes, and equipped with quick-reference scripts covering multiple use cases.
• Notable Statistic: They have 12 AI use cases live, all with robust testing and member communication.
• Quote:
  “Letting their staff realize that we're augmenting them, we're not replacing them, that they are part of the solution...”
  — Betsy Williamson, 03:34

4. Member Acceptance & Satisfaction with AI (04:55–06:48)

• Contrary to expectations, members were receptive to AI calls.
• Key Statistics:
  • Stephanie’s satisfaction rate among members: 9 out of 10
  • 93% rated Stephanie a 7 or higher in satisfaction
  • Engagement and call completion rates matched or exceeded those with human agents
• AI’s unlimited “handle time” increases rapport and support; Stephanie excels in motivational interviewing and identifying complex member needs.
• Quote:
  “We did see similar acceptance rates with humans versus our AI agent and satisfaction rates with the call.”
  — Betsy Williamson, 05:29
  “She is probably the greatest expert I've ever seen in motivational interviewing.”
  — Betsy Williamson, 06:14

5. Impact on Care Team Workload & Satisfaction (06:48–08:18)

• Care team capacity expanded by 360%
• Stephanie resolved 93% of calls in a single interaction; 6.7% needed escalation to human care (typically cases with low health confidence).
• Human care teams now focus on high-need, complex cases—improving both staff and member satisfaction.
• Quote:
  “We saw 360% capacity within our care teams, which really allowed them to work with those members that needed their help.”
  — Betsy Williamson, 07:06

6. Real-World Application: Diabetes Management Program (08:18–09:29)

• In the chronic condition management program for diabetes, follow-up calls by Stephanie uncovered “invisible” needs:
  • 19.1% of members disclosed issues like anxiety, loneliness, or financial hardship (signals not visible on claims data).
  • Medical Mutual could proactively connect members to needed community resources.
• Quote:
  “19.1% of members... actually disclosed some things like anxiety, loneliness, or even financial hardships, which were all signals that are invisible on a claim.”
  — Betsy Williamson, 08:53

7. Looking Ahead: The Future of AI in Population Health (09:29–10:20)

• Medical Mutual sees infinite potential AI use cases to deepen member engagement, allow for earlier intervention, and improve outcomes.
• Vision: continuous member engagement, early intervention, and strengthening relationships to reduce costs and improve health.
• Quote:
  “Our vision right now is really continuous member engagement and early intervention and strong relationships.”
  — Betsy Williamson, 10:11

8. Advice for Other Healthcare Leaders Considering AI (10:20–10:51)

• Choose the right partner, especially for clinically sensitive applications—safety must come first.
• Build trust with both your partner and your internal team; transparency removes fear and stigma around job loss.
• Implement a thorough change management plan for stakeholder buy-in.
• Quote:
  “The more transparent you are with the teams, you're able to remove the fear and the stigma that they're not going to have jobs. And then, of course, a great change management plan.”
  — Betsy Williamson, 10:43

---

Notable Quotes & Memorable Moments

• “The train's already left the station, right? You're going to either be on the train or you're going to be on the platform.”
  — Betsy Williamson, 03:12

• “We did see similar acceptance rates with humans versus our AI agent and satisfaction rates with the call. We got 9 out of 10 satisfaction rate with Stephanie...”
  — Betsy Williamson, 05:29

• “She's not held to call center metrics, so she doesn't have an average handle time. So she's able to make those relationships. And she is probably the greatest expert I've ever seen in motivational interviewing.”
  — Betsy Williamson, 06:09

• “We saw 360% capacity within our care teams, which really allowed them to work with those members that needed their help.”
  — Betsy Williamson, 07:06

---

Major Timestamps

• 01:12: What makes Medical Mutual different from national carriers
• 02:04: Decision to implement an AI agent (“Stephanie”)
• 03:09: Change management and internal staff buy-in
• 05:09: Member acceptance, satisfaction, and engagement with AI
• 07:03: Effects on workload and focus of care teams
• 08:29: Case study: Diabetes management program outcomes
• 09:42: The future vision and next steps for Medical Mutual’s AI implementation
• 10:27: Betsy's advice for other payer executives considering AI

---

Tone and Language

The discussion is candid, practical, and optimistic, with a clear focus on transparency, member-centricity, and responsible innovation. Williamson’s language is direct and community-focused, repeatedly reinforcing Medical Mutual’s role as a “hometown insurer” and the importance of scaling with empathy and safety.

---

Useful for Listeners Who Haven’t Tuned In

This episode provides a comprehensive look at how a regional insurer can safely and ethically implement AI to benefit both members and staff, offering actionable insights, hands-on experiences, and practical guidance for healthcare leaders exploring similar paths.

Transcript

[A]

Generative AI is driving the era of infinite healthcare abundance. At Hippocratic AI, we partner with more than 60 innovative healthcare organizations across payers, providers and life sciences to drive patient and member outcomes, offering the safest clinical AI voice agents. Hippocratic AI is the first company to define comprehensive voice AI benchmarks for healthcare safety, clinical accuracy, empathy, consistency. Learn more@hippocratic AI.com hi everyone, this is Grace Lynn Keller with Becker's Healthcare. Thank you so much for tuning in to the Becker's Healthcare podcast series. Joining me for today's discussion is Betsy Williamson, R.N.B.S, mHA, Vice President of Quality Performance and Population Health at Medical Mutual of Ohio. So Betsy, thank you so much for being here today.

[B]

Good morning, thank you for having me.

[A]

And let's dive right into this conversation. So you lead Quality Performance and Population Health. One of the oldest health insurers in the country. Medical Mutual of Ohio was founded in 1934 and serves over 1.2 million members. For people who don't know Medical Mutual, what makes the organization different from the larger national carriers?

[B]

Yeah, that's a really great question and one we really like to talk about. So Medical Mutual, we're known as the hometown insurer. We are a mutual company, which means that we are owned by our members and, and not by Wall Street. And that really gives us the ability to focus on our members, our employees and our communities, which we do a lot with. We live in the same neighborhoods as our members do, we use the same providers, we are shopping in the same stores. So we really understand our members needs and what is happening within the community which allows us to really care for our members in a better way.

[A]

And one of Medical Mutual's brand promises was you call us, you get a human. And that's almost the opposite of what most people think AI is about. So what made you decide to bring an AI healthcare agent in despite that ethos?

[B]

So we recognize that we do a really great job. And you're right, we are known for our member services and our care team and working directly with those members. But we also know that when we can work together with those members that we can make a a measurable impact in their lives to help them be healthier in whatever stage that is. And so in order to do that, we really need to be able to scale it. And not everybody needs to speak directly to a live human, but they do need to make that connection and be able to get directed into a live human when there is a need there. So I think realizing that Stephanie does not replace humans And Stephanie is the name of our AI agents, but she actually augments it. So, you know, we're able to expand more and reach more people and help them through their journey.

[A]

And every payer executive in this room is dealing with some version of internal resistance to AI. So how do you get your own team, especially frontline staff and leadership, to be comfortable with this?

[B]

Well, Grace, honestly, the train's already left the station, right? You're going to either be on the train or you're going to be on the platform. And I think that it's important to recognize that one in three people today are using some type of AI to explore healthcare decisions, health care problems. And so if we can intervene in that piece and be available to them to help them make the final decision, that is where we really need to be. So I think the big piece that everybody, as they approach this, really needs to worry about change management. So letting their staff realize that we're augmenting them, we're not replacing them, that they are part of the solution, that they understand the escalation process, the preparation that goes into it, the testing. You know, we were able to give our call center 30 second scripts for every single use case we have. And we have 12 use cases right now that are out there. And we also informed our members know anytime that Stephanie calls a member, she is in. She's letting them know that she is an AI agent that is part of Medical Mutual. So the members aware that they're talking to an AI agent and they also are aware that at any point they can touch a button and get a live person. We also, in addition to that, when we were first rolling it out, we sent out member flyers and let the members know that Stephanie was going to be calling them. And I think that was really the key that got us our highest completion rate was letting the members know and they're aware. So when they answer the call, they're willing to talk. So it is really a compliance first process. And we've had zero data breaches by doing that.

[A]

That's wonderful. And you launched Stephanie across 13 programs covering everything from chronic disease management to medication adherence to onboarding. And I'm curious what surprised you most about how members responded?

[B]

I think that the thought out there is that people don't want to talk to AI and that they're not going to answer the phone. And I think that we really saw that that's not the case. Members are embracing AI just the same as everyone else with whether it's ChatGPT or Copilot. On your phone or your devices or Alexa or any of the number of other ones that we have. But we did see similar acceptance rates with humans versus our AI agent and satisfaction rates with the call. We got 9 out of 10 satisfaction rate with Stephanie and that had 93% of those rated her a 7 or higher. So it was all very positive. And then we are seeing that Stephanie is able to take time. She's not held to call center metrics, so she doesn't have an average handle time. So she's able to make those relationships. And she is probably the greatest expert I've ever seen in motivational interviewing. So she's able to catch on those little key points that a member might, might throw out there from a customer service rep that they might miss that would trigger them to need additional help. So we've really found that she's been a great addition to our team. And we did some internal studies as well that showed that members find that she's able to explain difficult, complex situations like member benefits, sometimes better than our human agents are. So overall, I think it's been a very positive experience and that people want to engage. Probably the biggest thing that came forward.

[A]

Absolutely. And kind of on the flip side of that, there is a lot of fear that AI will replace jobs. So what actually happened to your care team's workload and focus once the AI healthcare voice agent was live?

[B]

So once the AI voice agent was live, we saw 360% capacity within our care teams, which really allowed them to work with those members that needed their help. We had 93% of the calls that Stephanie made were resolved with just one touch with Stephanie that didn't require human intervention, but there was 6.7 that did. And those were the ones that had that low health confidence. And so they were able to connect with our nurses, our physicians, our pharmacist, our care teams, customer service, and they were able to work through some of their concerns and get them where they were. You know, I think that this has really improved our employee satisfaction as well as member satisfaction because we're able to touch more. So the employees are feeling more connected to us as their health insurer. And it just goes back to driving home that we are home health or hometown insurer. But it also our employees are much more satisfied in the work that they do because they're just not cold calling people or they're not just taking calls that really doesn't work at the top of their license. So now they're really making a difference where they need to make A difference?

[A]

Absolutely. And I'd love to know, can you walk us through a specific program or moment that made this feel real for you where you thought, okay, like, this is actually working for us?

[B]

Yeah. So one of our first use cases was with our chronic condition management program for diabetes, and we require them to engage with a coach and get very educated on their diabetes program. And then once they graduate from that program, we still require them to call in on a quarterly basis and do a wellness check in on how they're doing, what their hemoglobin A1C is, and if they do that, then they get their diabetic supplies for free. And so through these calls, we had very high engagement. And we also found that there was 19.1% of members who actually disclosed some things like anxiety, loneliness, or even financial hardships, which were all signals that are invisible on a claim. And we were able to intervene and get some care out through there, find them resources in their community and help with that as well. So I think that's a really big win.

[A]

Absolutely. And so, looking forward, you've launched with several great use cases that you've already seen success with. So what does the next 12 months look like for Medical Mutual and your AI healthcare voice agent?

[B]

Well, I think that we're still brainstorming. The number of use cases, really, I believe, is infinite, and we want to look at how do we expand our full population. If you're able to deepen the population health and the condition management activities that you do with the members, you're able to really drive that relationship. So you will know earlier when there's a problem that's arising, which will really help with driving down medical costs and improving outcomes overall for the members. So, you know, our vision right now is really continuous member engagement and early intervention and strong relationships.

[A]

And if another payer executive came up to you after this podcast and said, said that they were interested but said they don't know how to get started, what would you tell them?

[B]

I would say take the time and pick the right partner. You need to have somebody, especially in a situation where it's clinical, that has the priority of safety first. And then I think you need to build trust with that partner and be completely transparent with your teams. The more transparent you are with the teams, you're able to remove the fear and the stigma that they're not going to have jobs. And then, of course, a great change management plan.

[A]

Well, Betsy, thank you so much for sharing your time and these insights today and you can tune into more Becker's healthcare podcasts by visiting our podcast [e-mail] thank you.