Patient communication automation
Lab reports delivered, explained and queried without a call centre
A diagnostic chain processed 20,000 reports a day and ran a 60-seat call centre to tell patients their report was ready and answer “is this value normal”. We built a WhatsApp workflow that delivers reports securely, explains results in plain language within safe boundaries, routes clinical questions to pathologists, and books follow-up tests.
- Client
- Diagnostic laboratory chain
- Region
- India
- Scale
- 140 collection centres, 1,800 employees
- Timeline
- 11 weeks including clinical review of all explanation templates
Representative engagement. Client details are anonymised and figures reflect typical outcomes from comparable deployments.
What the client got
0%
Reduction in inbound call volume
Report status and basic explanation questions no longer reach the call centre.
0.0/5
Patient rating of report explanations
From in-chat feedback across 200,000 reports.
0%
Of recommended follow-up tests booked
Up from an estimated 8 percent with no reminder.
The bottlenecks
Bottleneck 1
Report-ready calls consumed the call centre
Most calls were patients asking whether the report was ready, followed by requests to email it again.
Bottleneck 2
Patients did not understand the report
A page of values and reference ranges generated anxious calls. Agents without clinical training gave inconsistent answers.
Bottleneck 3
Follow-up tests were lost
Doctors ordered repeat or confirmatory tests that patients forgot. There was no reminder and no easy way to book.
What we built
Step 1
Secure delivery
When a report is signed off in the lab information system, the patient receives a WhatsApp message with an OTP-protected link. Doctors who ordered the test receive their copy in parallel.
Step 2
Plain-language explanation within guardrails
An LLM explains what each test measures and whether values fall inside the reference range, in the patient’s language, using pathologist-approved wording. It does not diagnose, and any abnormal value includes a prompt to consult the ordering doctor.
Step 3
Clinical query routing
Questions beyond the guardrails are routed to the duty pathologist with the report attached, and the answer goes back through the same chat. Critical values trigger a direct call per existing protocol.
Step 4
Follow-up test booking
Recommended repeat tests trigger a reminder at the right interval with home collection booking in the chat.
The automation workflow
Report sign-off to patient understanding
Delivery, explanation and follow-up
Questions clients ask about this engagement
How do you stop the AI giving medical advice?
Explanations are generated from pathologist-approved templates per test, limited to what the test measures and where the value sits. Anything resembling a diagnosis or treatment question is routed to a pathologist.
What about patients without WhatsApp?
SMS with a secure link and the IVR remain available. Around 9 percent of patients use those channels.
More in Healthcare
Industry overview
Healthcare workflow automation
Patient intake and reminders that cut no-shows for a multi-clinic group
Front desk staff re-keyed paper intake forms, chased confirmations by phone, and still faced a double-digit no-show rate. We automated digital intake, insurance checks and multi-channel reminders with privacy controls built in, so clinicians see more patients and staff do less data entry.
38%
Reduction in no-shows
B2B lead generation
Reaching hospital CIOs for a healthcare IT vendor
Selling into hospitals means long cycles and cautious buyers who ignore generic outreach. We built an account-based outbound programme for a healthcare IT vendor that reached CIOs and department heads with specific, relevant messages and produced a steady flow of discovery meetings.
12
Discovery meetings per month
Patient recall and scheduling automation
Hygiene recalls and no-shows handled by workflow across 18 dental practices
A dental group’s hygiene schedule was the engine of its revenue, and a fifth of slots went empty. Recall lists were printed, front desks called when they had time, and no-shows were discovered when the chair sat idle. We built a recall and confirmation workflow that fills the schedule from the practice management system and works the waitlist automatically.
21%→7%
Hygiene no-show rate
B2B lead generation for medical devices
Reaching procurement and department heads for a medical device distributor
A distributor with new lines in patient monitoring and infection control relied on tenders and relationships with a dozen hospitals. Expansion to private hospital groups and day-surgery centres needed conversations with biomedical engineers, department heads and procurement before tenders were written. We built an outbound engine around new facility openings, accreditation cycles and equipment age.
47
Clinical evaluations booked
Have a similar bottleneck?
Tell us how the work happens today. We will show you the workflow we would build and prototype it on your own process.