By implementing a VoiceAI Agent, Clearwater Benefits automates verifications and reduces handle times while managing large call volumes.
Navigating health insurance is inherently complex, often requiring callers to share detailed demographic and eligibility information before issues can be resolved. Clearwater Benefits focuses on simplifying that experience while maintaining the rigor required in healthcare administration.
The company operates contact centers both in the United States and offshore, supporting a broad and growing population of members and providers. As volumes increased into the thousands of daily calls, Clearwater identified intake as a critical lever for modernization.
“Verification alone takes a minute or two of our agents’ time,” explains Patrick Kolean, Senior Manager of Technology at Clearwater. “They’re looking up details and doing a lot of manual work before they can even start helping the caller.”
That manual effort compounded across calls, pushing average handle time to roughly eight minutes and limiting agents' ability to focus on higher-value work.
At the same time, Clearwater was transitioning core systems, moving from Zendesk to Vonage for telephony and from Google Sheets to Salesforce for customer information. This created an opportunity to rethink intake and routing with a more scalable, integrated approach.
Clearwater explored automation solutions that could streamline its verification process and ease pressure on frontline agents.
“Our CEO found Observe.AI and recommended it for the post-call QA work,” shares Kolean. “That was our initial interest. We were then shown the AI agent while we were starting to look at moving to another telephony provider.”
Clearwater eventually adopted Observe.AI’s VoiceAI Agent, naming the agent Amy.
“We were just really impressed with the system’s flexibility,” adds Kolean. “From what we saw, the build was very AI-driven. You just write out what you want it to say or do, instead of building a rigid flow of rules.”
Before adopting VoiceAI Agent, Clearwater used Zendesk’s interactive voice response (IVR) system. This setup didn’t provide a way to automatically adjust call routing during public holidays. Staff had to manually change the IVR each time to prevent callers from being routed incorrectly or receiving outdated prompts. This added unnecessary and difficult work, with the potential for increased error risk.
There were also routing inefficiencies. The three-layer IVR menu funneled most callers into just two groups, which increased wait times. As Kolean points out, “Time in the IVR was way more than it needed to be.” The experience did not match the level of quality and personalization that Clearwater Benefits set out to provide.
Clearwater prioritized dynamic, data-driven call flows that could adapt automatically and reduce unnecessary time in menus. There was also an emphasis on the necessity for the AI agent to have a natural-sounding voice, to maximise acceptance among callers.
“The conversation and fluidity of Observe.AI’s agent sounded like a real person,” he says. Such natural delivery helps avoid the resistance callers sometimes feel with automated systems.
Concerns about AI adoption were minimal. While Kolean acknowledges that some members might prefer speaking directly to a human, he highlights the cost benefits. “Automating routine intake allows our agents to focus on higher-value interactions that require judgment and expertise.”
The VoiceAI Agent, named Amy, launched with member authentication and quickly expanded to support provider verification as well. Today, Amy handles verification in both English and Spanish, collecting required details such as name, phone number, and partial Social Security numbers.
If an initial verification attempt fails, the agent initiates a second pass to gather additional information before transferring to a human agent. Callers can request a live agent at any point, though that occurs only a small number of times per day.
During peak periods, Amy absorbs large volumes of short verification-only calls, while escalations are transferred with full context to human agents. This prevents callers from repeating information and allows agents to begin each conversation already informed.
The VoiceAI Agent also manages tagging, routing, and alerting automatically, removing manual steps for both agents and supervisors.
With verification automated, Clearwater has seen measurable improvements in intake efficiency. Verification now averages ~90 seconds, while overall handle time has dropped to ~5.5 minutes, down from roughly eight minutes.
Member verification rates have stabilized between 80–85%, with provider authentication continuing to improve as the system learns from real interactions.
While the per-call time savings may appear incremental, applying them across thousands of weekly interactions has materially increased agent capacity. Human agents spend less time on repetitive intake tasks and more time resolving eligibility questions, benefits issues, and care coordination needs.
Kolean reports strong ROI since deployment, driven by the AI Agent’s ability to streamline authentication and routing while taking that extra workload off human agents. The lift in the customer journey and the day-to-day experience for agents has been consistently positive.
Rather than focusing on narrow cost savings, Clearwater evaluates success by reducing human-handling time, increasing throughput, and improving operational resilience, supporting sustainable growth as membership and provider volume increase.
“There’s no doubt implementing Amy has materially increased our operational capacity and improved how we scale our contact center,” says Kolean. “It’s taken a significant load off our team.”
“There’s no doubt implementing Amy has materially increased our operational capacity and improved how we scale our contact center.”
— Patrick Kolean, Senior Manager of Technology, Clearwater Benefits
An unexpected, but welcomed, benefit has been a decline in callback requests. Kolean notes that on a day when Amy was temporarily offline, callers were more likely to abandon or request callbacks. With the AI active, callers stay on the line instead of opting out.
“They might wait a few more minutes because they’ve already had a conversation (with Amy, the AI agent). It eases the load on the call center rather than having everything come in at once.”Navigating health insurance is inherently complex, often requiring callers to share detailed demographic and eligibility information before issues can be resolved. Clearwater Benefits focuses on simplifying that experience while maintaining the rigor required in healthcare administration.
While implementation initially focused on verification, Clearwater will expand automation to address other inquiries such as provider information, network FAQs, and claims status updates, and to give agents relevant context before transferring a call.
Clearwater views this automation layer as foundational infrastructure, designed to support future expansion across eligibility, claims status, provider navigation, and member education.
The company is also considering how to enhance accuracy, expand classification conditions for Spanish-speaking callers and unresponsive interactions, and improve how performance metrics are presented.
To support future automation initiatives, Clearwater is creating a more unified dataset through Salesforce. Kolean doesn’t see this as a major transformation. To him, it’s a logical next step and gradual expansion of what already works by using AI to handle routine interactions so support agents can spend more time resolving benefits issues and coordinating care.
“We’re taking it one step at a time,” says Kolean. “The automation we have right now already takes a big load off our team. And as we expand it with the support of Observe.AI, we just expect more of those routine calls to be handled up front.”
“The automation we have right now already takes a big load off our team. And as we expand it with the support of Observe.AI, we just expect more of those routine calls to be handled up front.”
— Patrick Kolean, Senior Manager of Technology, Clearwater Benefits
.jpg)
