Stellarus Launches Real-Time AI Copilot for Health Plan Agents

What happened

Stellarus has launched a real-time AI copilot designed for health plan customer service representatives, built to support live member interactions as they happen. The tool is positioned to reduce cognitive load on agents and lower the risk of misinformation being passed to members during calls.

Rather than replacing the human representative, the copilot works alongside them, surfacing guidance in real time as a call unfolds. The launch reflects a wider push in health insurance customer service toward tools that assist agents at the point of interaction rather than only after the fact.

Why it matters

Health plan customer service is a high-stakes, high-complexity environment: representatives must navigate benefit rules, eligibility details and policy nuances in real time, often while a member is anxious or confused about coverage. Errors or inconsistent answers in this setting carry real consequences, from member distrust to costly claims disputes. A copilot that reduces the mental burden on agents mid-call addresses a genuine service-design pain point rather than a peripheral inefficiency.

From a behavioral-economics lens, this is as much about the representative's decision environment as the member's experience. Agents under cognitive load default to shortcuts, guesswork or inconsistent scripts — precisely where misinformation creeps in. Tools that lighten that load at the moment of decision, rather than through after-call coaching or retrospective quality checks, target the root cause rather than the symptom.

The Renascence take

The interesting part of this story isn't the AI itself — it's where Renascence sees the intervention placed: inside the live interaction, not around it.

Most contact-centre AI investment still targets what happens before or after the call — routing, summarisation, sentiment scoring. Very little targets the exact moment an agent is deciding what to say next, which is where trust is actually won or lost. A copilot embedded in the live conversation is a bet that reducing an agent's cognitive load, not just their handle time, is the real lever for consistency and member trust. The operators who get the most value from tools like this won't be the ones who buy the software — they'll be the ones who redesign agent workflows and incentives around it, so representatives are rewarded for accuracy and calm judgement, not just speed.