Healthcare / clinic · Dubai · Al Barsha + Jumeirah · AI Enablement
How a DHA clinic automated 71% of patient enquiries without losing the human touch
71% of inbound patient enquiries automated. Zero DHA audit findings. First-response time from 4h → 9 seconds.
Client: A DHA-licensed multi-specialty clinic (Al Barsha + Jumeirah)
71%
Enquiries automated
with human handover on the 29%
from ~4h → 9 seconds
First-response time on WhatsApp
from 22% → 8%
No-show rate
zero
DHA audit findings
first post-deployment audit
The challenge
- 300+ inbound patient WhatsApp messages per day across two locations with two admin staff triaging.
- First-response time on WhatsApp averaged 4 hours during clinic hours, 12+ hours overnight.
- 22% appointment no-show rate — reminders sent manually via SMS after the fact.
- Facing an upcoming DHA audit with concerns about data residency and consent capture.
What we built
- Deployed Solinify Pulse with an AI patient-flow agent tuned on the clinic's specialty menu and doctor schedules.
- WhatsApp booking with real-time doctor availability + Arabic + English booking flows.
- Automated appointment reminders (24h + 2h before) via WhatsApp with one-tap reschedule.
- Migrated patient records to UAE-hosted AWS Middle East region ahead of DHA audit.
- Consent capture flow with timestamped audit trail — meets DHA Health Data Protection Regulation.
What actually happened
- 71% of enquiries handled end-to-end by the AI agent (booking, rescheduling, FAQ, directions).
- First-response time dropped to 9 seconds median — even overnight and on weekends.
- No-show rate dropped from 22% to 8% inside the first 2 months of automated reminders.
- First DHA audit post-deployment closed with zero findings.
- Admin staff redeployed from message triage to patient experience and insurance claim work.
Tech stack
- · Solinify Pulse
- · WhatsApp Business API
- · FHIR R4 data model
- · SNOMED CT
- · AWS Middle East
- · eClaim integration
Timeline
6 weeks from kickoff to go-live
“The AI agent handles what needs speed. Our team handles what needs empathy. That split is what made this actually work — not the technology, the operating model around it.”