Sutter Health Joins Health Systems Bringing Agentic AI Into the Call Center
Key Highlights
- Hyro says its conversational AI can resolve up to 85% of routine patient interactions, including appointment management and billing inquiries.
- Sutter Health’s contact center employs approximately 1,750 FTEs focused on patient engagement.
- The deployment of Hyro’s platform aims to improve patient experience by enabling quick, natural language interactions across multiple channels like chat, texting, and live support.
One of the areas where health systems are innovating with agentic AI is in the call center. For example, California-based Sutter Health is partnering with a company called Hyro, whose agents can help patients with appointment management, scheduling, prescription management, billing inquiries and more. Healthcare Innovation recently spoke with Steve Chambers, the nonprofit integrated health system’s vice president of enterprise contact center, about the deployment.
Hyro has strong ties to the healthcare sector. On Oct. 21, the company announced $45 million in new growth funding led by Healthier Capital, with participation from Norwest and Define Ventures, as well as other existing investors. The round also included new strategic investments from Bon Secours Mercy Health, one of Hyro’s long-standing clients, and ServiceNow Ventures, the investment arm of ServiceNow. The financing comes just 10 months after Hyro’s previous raise and doubles the company’s valuation, bringing total funding to $95 million.
Hyro says it combines the flexibility of Large Language Models (LLMs) with its proprietary conversational engine, including its Small Language Models (SLMs) for healthcare organizations, and advanced knowledge graphs built for healthcare. The company adds that this hybrid architecture enables its AI agents to accurately resolve up to 85% of routine patient interactions.
Besides Sutter, Hyro’s platform is already deployed across more than 45 health systems, including Tampa General Hospital, Prisma Health and Piedmont Healthcare.
Healthcare Innovation: Steve, could you first describe your role at Sutter and the size of the operation that you oversee?
Chambers: The contact center plays an important role in Sutter Health. We have about 1,750 FTEs that are taking calls and connecting those patients to various points within the system. The large majority of those are inbound calls for scheduling patients. I like to call it anything you need to get from your doctor's office. We say scheduling, but it oftentimes refers to appointing, general directions, medication refills, and then billing, both ambulatory and acute. We have a health plan that we support as well as I'll call it tier one help desk support for our patient portal. It's a lot of password resets and site navigation.
HCI: Could you talk a little bit about the earlier stages of technology in trying to automate contact center work and perhaps some of the shortcomings that those solutions had?
Chambers: When I think of that phone call into a service center, for years and years, it has been press one for this or press two for that. And that's very tricky to automate. It's a very long process. If you're calling in and we say, Now please state your name. I got Steve. Is that correct? Press one, if yes, press No, if two. When you think of a patient experience, there are few times you're excited to call your doctor's office. Most times we don't feel well when we call. And so the greatest respect that we can have for our patients is to respect their time, as well as, of course, keeping them safe and protecting their data.
There have been a lot of advances in conversational AI, and our patients have had these experiences in other industries — airlines and banking — where they get to speak naturally with their phone system and be routed appropriately, and we really want to bring that same technology to our patients and respect their time in that way.
HCI: As you looked for a solution partner in the AI space, were there some priorities or requirements as far as interoperability with other tech you already have in place, or other things that were must-haves with a solution?
Chambers: There are a couple of problems that we're really trying to solve by bringing Hyro in. You mentioned interoperability, and being able to connect to our EHR, being able to read back information is important. We want to make sure that we've got multiple channels, that we're meeting a patient where they want to be met, whether that's chat, whether that's live agent, two-way texting. Hyro helps us do that.
HCI: Did you have specific objectives you were hoping to achieve, as far as timeliness?
Chambers: We want to reduce friction for our patients. We want to make sure that we're being respectful of their time. We want to make sure we're answering in a very quick way. This allows us to give a 24-hour self-service channel. That's important for a patient, those 2 a.m. calls and your kid has an earache, and you can get in and get scheduled right away, those are the types of things. We are reducing friction through these advanced tools, authenticating our patients quickly, and really respecting their time. And then, as I mentioned, being at the forefront of the omni-channel experience in healthcare is really important to us.
HCI: How did you hear about and test Hyro’s agents? Were you able to see it in action at other health systems or get references from other health systems? And is there a lot of competition amongst vendors in this space? Are there a lot of choices available that have a healthcare focus?
Chambers: This conversational AI has advanced quite a bit in the contact center industry as a whole. Certainly, there are several players out there. Anytime that we look at this, we look at how a potential vendor can fit into the whole suite, how it can help us meet our objectives. So we went through a robust process, and Hyro ticked the boxes on the things that were really important to us.
HCI: Was experience in the healthcare industry one of those? Did that give a certain level of comfort seeing they work with other healthcare systems?
Chambers: Yes, experience in the healthcare system was certainly important to us.
HCI: Do Hyro’s AI agents automate routine interactions, leaving the customer support reps to handle more complex ones? And if so, can you give examples of the things it can can handle vs. things that might require a handoff for a human intervention?
Chambers: At first, the types of things that we anticipate Hyro handling are some of the lower-hanging fruit, the easier questions. That might be a password reset on the portal; that might be general directions. Those are quick questions that can be answered. What our agents really love to do is take care of people and help them through their needs. Healthcare is tricky, and we have patients with very complex needs, as any healthcare system does, so this will free up time for the agents to do what they really excel at, which is helping patients through those complex needs, give them the empathy, spending time with them on the phone and getting them connected to their care providers.
HCI: Is one of your metrics of success how often the customers successfully complete tasks in their interaction with Hyro versus hanging up or asking to speak to a to a human?
Chambers: We'll certainly want to look at that more from a patient experience aspect, because that could be indicative that we’re not making it as easy for patients as we thought we were. We'll look at how many calls we are containing. How's the accuracy? How’s the patient experience? And we'll iterate, of course, as we go. Is there an opportunity for us to add to the things that Hyro could answer?
About the Author

David Raths
David Raths is a Contributing Senior Editor for Healthcare Innovation, focusing on clinical informatics, learning health systems and value-based care transformation. He has been interviewing health system CIOs and CMIOs since 2006.
Follow him on Twitter @DavidRaths
