How 3 Healthcare Leaders Use Agentic AI to Personalize Healthcare

If ordering a Burger King Whopper is easy and customizable, shouldn’t basic healthcare tasks be simple, too? AI agents just might be able to get us there.

// By Wendy Margolin //

In 2026, when consumers can get deliveries on demand and plan custom vacations with AI, a personalized patient healthcare journey seems like a baseline expectation — not a future aspiration.

Andrew Chang, chief marketing officer, UChicago Medicine

Agentic AI has the potential to help health systems deliver personalized patient experiences that are commonplace in industries like retail and hospitality.

For example, consider the simple act of ordering a Whopper at Burger King, as Andrew Chang, chief marketing officer of UChicago Medicine, did recently in a webinar during AI in Healthcare Marketing Week.

If you’re craving a Whopper, you can walk in, drive through, order online, have it delivered to your car curbside, or choose delivery. You can change your mind halfway through and decide to go through the drive-through instead. You can speak to a human or not. You can also customize every part of your Whopper.

Compare that experience to booking a simple healthcare appointment or even trying to refill your prescription. Many times, you can get help only during specific hours when a human is available.

Read on to see how leaders from UChicago Medicine, Northwell Health, and Modular Feedback use AI-enabled workflows to reduce friction, improve access, and scale communication.

Imagine healthcare where:

• Patient communication is tailored to your age, health history, and personal concerns.
• You can schedule an online appointment with the perfect provider for you at any hour of the day.
• You can have two-way communication with a human or AI agent and get assistance with your health or appointment-related needs.

But today, the patient journey still breaks down in familiar places like finding the right care, getting scheduled, receiving the right message at the right time, and having to repeat yourself. Basic personalization that’s so common in consumer apps remains elusive in healthcare.

Behind the scenes, healthcare marketers know just how much coordination, compliance, and technical complexity go into a personalized patient experience. The goal, says Chang, is to “find every single point of friction in a patient’s journey and put some WD-40 on it in the form of agentic AI to make it a lot smoother.”

Case Study 1: UChicago Medicine Uses Agentic AI to Reduce Access Friction Across Channels

When it comes to marketing, Chang states the obvious: Patients don’t judge your brand by what you claim, they judge it by what it feels like to get care. A bad patient experience can unravel all of your marketing efforts. Every touchpoint is part of your marketing communications, including ads, the website, the phone experience, text reminders, and handoffs.

AI strategy can get complicated quickly, but Chang says the goal is simple. “What we’re doing to elevate the human experience overall is to use AI to reduce friction,” he says.

Agentic AI has the potential to personalize this complicated patient journey at scale.

The marketing team began its patient personalization project by identifying and mapping every touchpoint in the patient experience. This included:

• Call centers
• MyChart
• SMS
• Emails
• Human phone calls
• Print mail
• Press Ganey surveys
• Billing

UChicago Medicine’s focus on the patient journey addresses the personalization challenge most systems face: Patients want to engage on their terms. Chang pushes back on the “digital front door” concept.

While some people want to engage digitally with their care team, others want to call a provider’s office. Chang stresses, “Digital front door is a terrible term that paints a picture in everyone’s mind that everyone has to go through a front door that’s digital. If you look at your patient base, most of them are probably Medicare age, over 65, and they don’t want to use digital, so why are we forcing them to go through one door that’s digital?”

Instead, he recommends healthcare organizations offer the Burger King approach of “have it your way.” This includes:

• Easy phone access for patients who prefer to call
• Text messaging for those who prefer SMS
• 24/7 access where possible, with seamless handoffs so patients aren’t forced to repeat themselves

Case Study 2: SCAN Health Plan Fixes Provider Accuracy

Chris Hemphill, founder and head of applied data science and AI at Modular Feedback

Chris Hemphill, founder and head of applied data science and AI at Modular Feedback, shares a use case that shows how using AI for “back office” work can directly change the patient experience. According to the “MIT State of GenAI in Business 2025,” 50 percent of GenAI budgets are allocated to sales and marketing, but back-office automation often delivers a better ROI.

Hemphill described how SCAN Health Plan, a non-profit Medicare Advantage program, used back-office automation to solve a practical problem. The organization lists provider information so users can check if the plan covers their providers. This means keeping tens of thousands of providers’ information constantly up to date. Verifying provider information requires manual browsing and cross-checking across thousands of records.

The work is tedious, but the stakes are high. “It might sound like I’m obsessed with spreadsheets and data, but I translate it to an 84-year-old woman who might not have access to an endocrinologist within a 30-minute drive,” says Hemphill.

With SCAN’s new system, outreach team members have thousands of agents browse their provider website, compare against source-of-truth data sources, and report discrepancies. This frees up thousands of hours to focus on action rather than research.

In the SCAN use case, ensuring the network and data are accurate improves the member experience. The case study applies to every large healthcare organization where accurate provider data is a prerequisite for personalization. If your directory is wrong, you can’t reliably tailor pathways, make recommendations, or provide appointment options. Fixing data quality is what makes “right care, right place, right time” realistic.

Case Study 3: Northwell Health Scales Personalized Communications

Paul Lambson, assistant vice president of consumer technology, Northwell Health

Paul Lambson, assistant vice president of consumer technology at Northwell Health, shares a marketing communications workflow designed to increase speed and consistency: an internal AI tool called CoStar.

As one of the largest health systems in New York, the content team was flooded with requests for web content, email, and printed materials. They needed support for first drafts so they could get the final copy out the door faster. To do that, Northwell built a workflow that generates early drafts based on the organization’s brand voice and existing content examples.

Today, the AI system helps with content such as:

• Subject lines
• Social posts
• Newsletters
• Patient communication templates

Here are the reported results from the pilot:

• 17 percent reduction in task completion time
• Tackled about 2,000 tasks per year
• Saved roughly 10 minutes per task

“The feedback we got from our copywriters was that this was a dream come true,” says Lambson.

When creative teams move faster, they can produce more personalized variants by tailoring messages to each audience segment. The AI agent helps Northwell maintain consistent communications across touchpoints without burning out staff.

Lambson recommends testing and measuring to determine what works best for your team. “We all see there’s a lot of value to agentic AI. It comes down to showing the value for these use cases by measuring the benefit you can provide to your company.”

Across three very different organizations, it’s clear the opportunities for agentic AI in healthcare are massive. Choosing the right use cases and the best ways to apply these tools will continue to be a learning process for all involved.