Kimber Molke, the director of the Nurse Health Line at Memorial Hermann Health System, faced the overwhelming task of having her agents make more than 5,000 follow-up patient calls each month. This January, she decided to try something new — an AI assistant named Grace. 

“We started using Grace in January,” Molke says. “We were spending an enormous amount of time and resources on follow-up calls. AI seemed like a potential solution.”

But could a robot truly replace the human connection of a follow-up call? Molke said her team and nurses were initially skeptical.

“Many people said, ‘I don’t want to talk to a robot,'” Molke says. “But Grace introduces herself as an automated assistant, and we made sure the call script sounded natural.”

Molke and her staff placed test calls to Grace, acting as “patients” to experience the process firsthand. The goal was to make the interaction helpful and efficient.

“We realized the devil was in the details,” Molke says. “We limited the number of times Grace calls someone, adjusted her language for clarity and worked on a smooth transfer process to a live nurse when needed.”

This attention to detail paid off. Grace quickly took over routine follow-ups, asking patients about their well-being, collecting survey responses, and seamlessly connecting those who needed additional care to Molke’s nursing team.

Exploring AI’s benefits 

The results have been impressive. Grace not only saves Kim’s team time but has also improved staff morale. 

“I had multiple community health workers making follow-ups,” Molke says. “With Grace, some of those workers are now available to take incoming calls, translate for our Spanish-speaking patients along with helping callers with getting a medical home.” 

Molke was mindful of how introducing AI might impact her staff. So far, Grace has saved 1,136 hours in full time hours staffed and completed 24% of surveys, when the caller is reached. 

“We talked about implementation for months before Grace arrived,” Molke says. “We let staff listen to calls; experience it as a patient. Transparency and showing the positive impact were key.”

“This showcases the real potential of AI during times when you may not have enough staff or may not be able to reach everyone. AI can bridge that gap.”

Tips for contact center leaders 

For health care leaders considering an AI solution, Molke has practical advice:

• Be transparent: Involve your staff early. Let them “meet” the AI tool and understand how it will make their jobs better.
• User testing: Have everyone experience AI from the patient’s perspective.
• Emphasize the positives: Focus on the time saved, the ability to better serve patients and how it allows staff to work at the top of their roles.
• Expect refinement: The set-up process takes time and flexibility.

“Don’t be afraid to challenge the way things have always been done,” Molke says. “AI can be a powerful tool for efficiency and better patient care. Grace has proven herself as a valuable member of our team.”