Health systems have a massive problem with patient outreach. Care coordinators spend eight hours a day leaving voicemails that nobody returns. Even the most sophisticated care programs fail when patients won’t engage.
Two AI healthcare companies think they have a solution. PromptWell, a patient engagement platform, has partnered with Ellipsis Health to build robocalls that can read emotions and respond accordingly. The goal is to turn cold outreach into actual conversations that get patients enrolled in care programs.
How will it work?
The system has two parts working together:
- PromptWell’s AI analyzes patient data to predict when someone will actually answer their phone
- When the call connects, Ellipsis Health’s AI agent called “Sage” takes over the conversation
Sage uses what the companies call an “Empathy Engine.” The AI listens to vocal cues in real-time and changes its responses based on the patient’s emotional state. It’s trained on millions of actual clinical phone calls, so it can have conversations that feel natural instead of robotic.
“PromptWell is built to automate the ‘how’ of engagement so healthcare teams can focus on the ‘what’ of care delivery,” said Kaushik Viswanathan, CEO of PromptWell. “Every touchpoint becomes more than just a notification, it becomes a clinically meaningful interaction.”
Why does it matter?
Patient engagement is a huge bottleneck for healthcare systems. The traditional approach of mass calling at scheduled times wastes time and money. Most patients don’t answer, and those who do often hang up quickly.
This partnership tries to fix both problems:
- Better timing means more patients answer their phones
- Emotional intelligence keeps patients talking instead of hanging up
The companies say this approach can actually get patients to complete Health Risk Assessments and enroll in care management programs – things that usually require multiple failed attempts by human staff.
The context
Healthcare systems are under pressure to improve patient outcomes while controlling costs. Value-based care programs can help, but only if patients participate. Traditional outreach methods have low success rates, making it hard to justify the investment in these programs.
AI voice agents represent a potential middle ground between expensive human outreach and ineffective automated messages. Several companies are working on similar technology, but the focus on emotional intelligence sets this partnership apart.
The challenge will be whether patients accept AI agents for healthcare conversations. Trust is critical in healthcare, and many patients may prefer human contact for sensitive medical topics.