At UC San Diego Health, an early deployment of Amazon Connect Health is reducing call times by 25% and reshaping how the health system approaches patient access.
The health system implemented Amazon Connect, an AI tool designed to automate administrative tasks and improve patient access to care, after reaching the limits of its legacy call center vendor.
“Our existing vendor presented persistent challenges — not just operationally, but in our ability to innovate on top of the platform,” Joshua Glandorf, CIO of UC San Diego Health, told Becker’s. “When it came time for contract renewal, we made a deliberate decision to move toward a more modern, innovation-oriented platform.”
The health system’s first use case with Amazon Connect was patient authentication. UC San Diego Health deployed an AI-powered agent that verifies patient identity by integrating directly with its EHR. When a patient calls, the AI handles authentication upfront and then transfers the call to a human agent.
“This removes a manual step for staff and allows them to immediately focus on higher-value tasks like scheduling or answering questions,” Mr. Glandorf said.
The health system rolled out the function across its call centers in December.
“The impact has been meaningful — about a 25% reduction in call time, roughly one minute per call on a four-minute average,” Mr. Glandorf said.
In a March 5 news release, Amazon Web Services said the health system reported redirecting 630 hours per week from patient verification to patient assistance and reducing call abandonment rates by 30%, reaching up to 60% in some departments using Amazon Connect Health.
That efficiency gain translates into improved access, he added, noting the health system can answer more calls, reduce wait times and create a smoother experience.
Beyond patient authentication, UC San Diego Health is working with Amazon Web Services to align its roadmap with organizational priorities.
“We see an opportunity to expand access. Today, our call centers operate during standard business hours, which doesn’t always align with patient needs,” Mr. Glandorf said. “With this technology, we can move toward more flexible, always-on engagement. That includes not only handling calls outside traditional hours but also expanding into other channels like text, voicemail and email.”
Ultimately, the goal is to create a more personalized, patient-centered communication model rather than a one-size-fits-all approach.
While the early results have been significant, Mr. Glandorf emphasized AI deployment comes with operational challenges.
“There’s a lot of excitement around AI, but also some hesitation,” he said. “What we’re seeing is that the technology is absolutely powerful and transformative — but it’s not instantaneous.”
Implementing AI, he added, requires significant operational effort.
“To automate workflows, you have to fully map and understand them — every detail, assumption and policy,” he said. “That’s work that historically lived in people’s experience and intuition, and now it has to be explicitly defined.”
That level of rigor, according to Mr. Glandorf, can slow progress initially.
“It’s more complex than simply turning on a new capability,” he said. “My advice is to stay focused on specific organizational needs rather than trying to broadly prepare for AI. If you try to prepare for everything, you risk not making progress. It’s more effective to identify high-impact use cases and go deep.”
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
