Smarter, simpler pharmacy support with AI: An expert interview

Ryan Perkins: Sarah, let’s jump right in. AI is everywhere right now. But why is it so important to connect AI to the member experience?

Sarah Greenwell: AI only matters if it makes something better for the person on the other end. A member is not thinking, “I hope my pharmacy benefit experience uses AI.” They are thinking, “Can I get my medication? Can I understand what this costs? Can I figure out what I’m supposed to do next?”

So, for me, the connection between AI and member experience is really practical. Can we make things easier? Can we make the experience feel less confusing? Can we help people get the right answer faster?

Ryan: That is exactly where the strategy has to land. From the AI strategy side, we can talk about models and capabilities and automation, but the real question is: What problem are we solving? AI gives us a chance to take a lot of complexity that sits behind the scenes and turn it into something more useful, more timely and more actionable for members.

Sarah: Right. And I think that is the key. The member should not have to understand all the complexity underneath. The experience should do more of that work for them.

Ryan: So, from a broader perspective, what is the core member problem AI can help solve? And can you talk about some of the early capabilities your team is building?

Sarah: Sure. The simple answer is, health care asks the consumer to do too much. Members have to interpret letters, understand coverage, figure out their cost. They're checking their prescription status, calling for help. Sometimes they're repeating information, comparing their options, and sometimes they still leave without a clear answer.

A great example is the work that we're doing with Optum Assistant, where my team's responsible for translating those AI capabilities into a member experience that helps people:

• Understand their pharmacy benefits
• Manage their medications
• Get answers in a way that feels simple and connected

Ryan: And what AI can help with is connecting information that is often fragmented. A lot of the answers exist somewhere, but they may not be easy to find, easy to understand or available at the right moment.

Sarah: And that “right moment” part is everything. A member does not need an answer three steps later. They need it when they are trying to make a decision. Is this covered? Why did my cost change? Is there another option? What happens next? That is where AI can help us move from just responding to members to actually guiding them.

Ryan: That is a good distinction. It is not just answering a question. It is helping someone move through the experience with more confidence.

Sarah: A lot of people hear AI and immediately wonder if it is about replacing humans. What do you think about that?

Ryan: I understand why people ask that. But I do not think the best use of AI in health care is replacing people. I think it is reducing the work that does not need to be manual, so people can focus on the work that really requires judgment, empathy and problem-solving.

There are many areas where AI can help: 

• Summarizing information
• Routing someone to the right place
• Finding relevant answers
• Identifying patterns
• Helping an advocate prepare for a conversation

Sarah: That is exactly how I think about it too. For me, AI should be a consumer accelerator, not a human replacement strategy. It should help members get answers faster. It should help them understand their options faster. It should help them take action faster. But it should not remove human support from the moments where people need it most.

Ryan: Right. And sometimes AI can actually make the human interaction better.

Sarah: If AI can take some of the friction out of the process — the searching, the repeating, the status checking — then our people can spend more time actually helping. So, the goal is not to take humanity out of health care. The goal is to take friction out of health care.

Ryan: So, how do you keep AI from making the experience feel less human?

Sarah: The best AI experience is probably one where the member does not think about AI at all. They just think, “That was easier than I expected.” Or “I understood what to do.” Or “I did not have to start over.” That is the goal.

Ryan: I agree. AI should not become another layer people have to fight through.

Sarah: Exactly. If AI becomes a wall between the member and the help they need, that is a bad experience. If it keeps someone from reaching a person when they need one, that is a bad experience.

But if AI helps clarify the issue, brings context forward, explains the next step and creates a smoother handoff to a human when needed, then it can actually make the experience feel more human.

Ryan: That is an important point. AI is not the relationship. It supports the relationship.

Sarah: Yes. And it should make the relationship better, not more distant.

Ryan: Personalization is a word everyone uses. What does it actually mean for members?

Sarah: Personalization cannot just mean putting someone’s name at the top of a message. Real personalization means the agent understands the situation.

Why is the member here? Are they trying to fill a medication for the first time? Are they worried about cost? Are they trying to understand coverage? Are they trying to avoid a delay? Have they already called once? Are they caring for a child or a parent? Those details matter.

Ryan: And AI can help make that kind of context more usable. It can help understand intent, connect data and guide the next best action.

Sarah: Yes, but I think the key is that personalization has to feel useful. A generic answer may be technically correct, but still not helpful. A personalized experience says, “Given where you are right now, here is what matters, and here is what you can do next.”

Ryan: And it has to be responsible. We have to use data appropriately, protect privacy and make sure we are not crossing the line from helpful to intrusive.

Sarah: Yeah. The sweet spot is when the member feels understood, not watched.

Ryan: So, where do you see the first real opportunities for AI to improve the member experience?

Sarah: Honestly, I think some of the biggest opportunities are in the everyday moments.

• Where is my prescription?
• Why did my cost change? 
• Is this covered?
• Is there a lower-cost option?
• Can I use home delivery?
• What does this letter mean? 
• What do I do next?

Those may not sound flashy, but they are the moments that shape how members feel about the experience.

Ryan: That is a really important point. Sometimes people want to jump straight to the most advanced use cases, but there is tremendous value in solving common friction at scale.

Sarah: Exactly. If we can make those basic moments clearer and faster, that is meaningful. And then over time, the bigger opportunity is connecting those moments so the experience feels less fragmented. 

Ryan: That is where strategy and experience come together. We start with the places where AI can safely and clearly reduce friction, then build toward a more connected journey.

Ryan: Sarah, affordability is such a major concern for members. How can AI help there?

Sarah: Affordability is a great example because it’s not just about the final dollar amount. It is about whether the member understands what is happening and what options they have.

A member may not know there is a lower-cost alternative. They may not know whether home delivery could help. They may not know whether a different pharmacy option matters. They may not understand why something changed. AI can help bring the right information forward earlier. 

Ryan: And that is a place where context matters. The answer may depend on the member’s plan, medication, pharmacy, timing and available alternatives.

Sarah: Right. And the goal is not to overwhelm the member with every possible option. It’s to make the most relevant options clear at the right time. So instead of a member being surprised at the pharmacy counter, we should be asking: could we have helped them understand this sooner?

Ryan: That is where AI can help turn complexity into guidance.

Sarah: Exactly. Not more information for the sake of information. Better guidance. 

Ryan: Another concern is adherence. Can AI help with medication adherence or help members stay on therapy?

Sarah: Yes, but I think we have to be thoughtful about how we talk about this. Staying on therapy is not always about forgetting. Sometimes the issue is cost. Sometimes it is side effects. Sometimes it is access. Sometimes the member is confused. Sometimes life just gets in the way. AI can help us better understand what kind of barrier might be present and what kind of support may actually help.

Ryan: That is the difference between generic engagement and intelligent engagement.

Sarah: Exactly. For one member, a refill reminder may be enough. For another, the issue may be affordability. For another, the right next step may be connecting with a pharmacist or care resource. So I do not think the goal is more outreach. The goal is more relevant outreach. 

Ryan: And that is where responsible design matters. AI can help identify signals and suggest support, but we need clear guardrails for when human or clinical expertise needs to be involved.

Sarah: Yes. AI can help us get smarter about support, but it should not oversimplify the human reality behind why people do or do not stay on therapy.

Sarah: Ryan, what concerns you most if AI is implemented poorly?

Ryan: Trust is the biggest concern. If AI gives inconsistent, inaccurate or poorly governed answers, trust erodes quickly. In health care, we have to be especially disciplined. Privacy, security, bias, explainability, accountability are not side issues. They are central to doing this responsibly.

Sarah: From the member experience side, my biggest worry is AI becoming another barrier. If a member is trying to get help and the AI makes them feel trapped, dismissed or misunderstood, then we have made the experience worse.

Ryan: That is where governance and experience design have to work together. 

Sarah: Yes. Because an answer can be technically generated and still not be helpful. So, the test I come back to is: did this reduce effort? Did it build confidence? Did it help the member move forward? If not, we need to rethink it.

Ryan: So, Sarah, what does good AI actually look like for the member?

Sarah: Good looks like less effort. The member gets an answer faster. They understand what is happening. They know what to do next. They do not have to repeat themselves. They are not bounced between channels. The experience feels like it has some memory and some common sense.

Ryan: And from my side, good also means the experience is scalable, consistent, and responsible. It cannot just work in one narrow use case. It has to be something we can build on. 

Sarah: Yes. And if we do it really well, the member may never say, “That was great AI.”

They will just say, “That was easy.” 
Or, “I understood what to do.” 
Or, “I felt supported.”

That is the win.

Ryan: That is probably the best measure of success. The technology disappears into the experience.

Sarah: That is the goal. Can you talk about what is the bigger vision for AI and member experience? 

Ryan: The bigger vision is a more intelligent, connected healthcare experience. AI can help bring information together, anticipate needs, support better decisions and create more consistent experiences across channels.

But we have to do that with discipline. We need ambition, but we also need responsibility.

Sarah: And from the member side, the bigger vision is that people feel less alone in the system.

Healthcare can feel like a series of disconnected moments — a letter here, a call there, a pharmacy counter issue, a portal message, a benefit question. AI gives us an opportunity to connect those moments and make the experience feel more guided. 

Ryan: That is a good way to say it. Connected, but still personal.

Sarah: Exactly. Not automated for the sake of automation. Guided in a way that feels useful. For me, success is when the member feels, “I understand what is happening. I know what I can do next. And I feel supported.”

Sarah: Before we go, Ryan, what is the one thing you would want people to remember?

Ryan: I would say AI is a powerful tool, but it only matters if it creates real value. It has to improve the experience, support better decisions and be used responsibly.

Sarah: And I would say AI should help us take friction out of health care, not humanity out of health care. That is the whole point.

AI should help members get answers faster, understand choices more clearly and take action with more confidence. And it should help our people focus on the moments where human support matters most. 

Ryan: That is the balance.

Sarah: Yes. AI should make the experience feel easier, more proactive, and more connected — not colder, not more complicated and definitely not less human.