An inbound AI receptionist for a diagnostic centre — and structured data from every call
A warm inbound voice agent that handles appointments, home sample collection, branch information and test queries — and turns every conversation into a structured record the centre can actually report on.
Client: Vijaya Diagnostics · Hyderabad, Telangana · Status: live
The challenge
Diagnostic centre phone lines are busiest exactly when staff are with patients. Calls are missed, and the ones that are answered leave no data behind — there is no record of what callers asked for, which tests they wanted, or how many enquiries never converted into a booking.
What we built
We deployed an inbound agent that answers every call in a warm, unhurried register and handles the real work: booking appointments, arranging home collection, explaining health packages and directing callers to the right branch. After each call an analysis pass extracts a structured record — department, service requested, preferred branch, date and time, urgency, sentiment, call outcome and whether an appointment was actually confirmed — so the centre gets an operational dataset instead of a call log.
Engineering highlights
- Configurable use cases: the clinic can add or change what the agent handles from a dashboard, without a code change.
- Per-call extraction into a fixed schema covering pathology, radiology, health packages and cardiology.
- Outcome is recorded explicitly — appointment booked, enquiry only, callback requested or incomplete — so conversion can be measured rather than estimated.
- The extractor is instructed to return null rather than guess, so a report never contains a detail the caller did not actually give.
- Deliberately soft conversational register, with natural acknowledgements, because callers to a diagnostic centre are often anxious.
Technology
Services used
Want something like this?
Tell us your problem and we'll tell you exactly how AI solves it and what it costs.