Voice AI · Healthcare
A 24/7 AI care coordinator for an international medical tourism company
The problem
Patient enquiries from four regions were waiting hours for a human callback.
Context
An international medical tourism company receives enquiries from the US, the Middle East, Europe and India through its website, social media ads and WhatsApp. Leads went cold while they waited, out-of-hours enquiries sat until morning, and call outcomes lived in people’s heads instead of the CRM.
What I built
- An AI voice agent that answers inbound calls and places an outbound call moments after someone submits an enquiry form.
- Region-aware routing: live transfer to the right team during working hours, and a clear callback promise outside them.
- Every call, transcript and recording logged automatically against the lead in the company’s CRM.
- WhatsApp acknowledgements and follow-ups sent the moment a lead arrives.
- Safety rules built in: consent capture, emergency detection and handover to a human, and the agent always discloses that it is an AI.
- Bilingual: English and Hindi.
Trade-off worth naming
A voice agent that sounds natural on day one isn’t the same as one that’s reliable. We planned a tuning period with real calls before scaling, which delays the full rollout but avoids patients hearing a half-trained agent.