eHealth Gives Callers the Regal Treatment

eHealth, an Indianapolis-based private online health insurance marketplace specializing in Medicare Advantage, prescription drug plans, and Medigap, gets inundated with calls during the annual Medicare open enrollment period every fall. In the past this prompted the company to hire and train hundreds of agents in a matter of weeks.

Due to the volume spikes, prospects and customers could spend hours in the queue if they didn’t drop off first. “We couldn’t provide enough staff so that they could have zero wait time,” says Ketan Babaria, eHealth’s chief product, technology, and artificial intelligence officer.

The problem became unsustainable, and so eHealth decided it needed a solution that could scale seamlessly during peak volume, screening inbound calls, qualifying prospects in real time, and transferring qualified leads to licensed insurance agents without adding headcount.

It started looking for technology in 2024 and eventually selected Regal’s customer engagement platform to transform its customer service and enrollment experience using AI voice agents.

“The other ones that we looked at sounded very robotic, and there were a lot of delays,” Babaria says.

Babaria also liked how collaborative the Regal team was. “In a way, we co-built the solution. We had a bunch of special requests, and they were very accommodating. They helped us refine our thinking as well,” he recalls.

The Regal platform also had the flexibility eHealth needed, enabling it to customize the solution to best meet its very specific requirements.

eHealth and Regal also had to work together to ensure all the guardrails were in place to protect customer health data and other personal information. “As you can imagine, there are a lot of challenges with hallucinations. And it’s a very regulated industry. We wanted to make sure that everything worked as expected, that we were compliant, and that the system was very empathetic,” Babaria says.

The platform was ready for a limited rollout pilot in February 2025. “We were getting constant feedback from our customers during the pilot phase, and we leveraged a lot of those learnings to actually launch it,” Babaria says.

During the pilot, eHealth learned that while the AI agent, named Alice, was extremely efficient in answering customer queries, it lacked sufficient empathy.

“We modified the whole system to ensure that Alice was very empathetic and patient,” Babaria says.

The change in responses was dramatic. Rather than just providing answers, Alice now responds as an empathetic human would, using phrases like “I’m sorry for your loss” if a customer calls about a spouse’s death. Those calls are now handled automatically, with an API-triggered removal from outreach lists.

eHealth also refined prompt logic after hearing how callers spoke; made long disclaimer statements more concise and shifted them later in the call flow to reduce hang-ups; and introduced cold transfers during peak volume to shorten handle time while keeping AI call summaries rich with context.

With Regal’s AI, every conversation now generates structured insights automatically. Those insights flow into Regal Improve, Regal’s analytics engine. eHealth can now analyze language distribution, top reasons for calling, and performance trends across workflows.

Regal’s platform also gave eHealth the scalability and control it needed, allowing it to automate complex screening workflows while meeting rigorous security and compliance standards. Alice executes approved disclosures and statements exactly as written, with deterministic prompting and full conversational logging for auditability.

The Regal solution has proved to be very successful, with all calls now being answered. The automated system is also outselling its human counterparts, with a 27 percent higher purchase rate. And 77 percent of consumers rate calls with Alice as “exceptional.”

Since the early days of the deployment, eHealth and Regal have expanded Alice’s capabilities beyond shopping and enrollment telephone support; the AI agent is now assisting with post-enrollment and general service calls, and eHealth expects to continue improving on the system.

“We want to continue to make the system more human, more empathetic, and more aligned to what our customers need,” Babaria says. 

The Payoff

Since installing Regal’s AI Agent solution, eHealth has seen the following results:

• every call is answered immediately with zero wait times;
• 77 percent of calls are rated “exceptional”;
• AI Agent’s purchase rate is 27 percent higher than human screeners; and
• calls are shorter overall.