Skip to content
← All case studies

AI Receptionists · Private build · 3 min read

CareLine WhatsApp Assistant

A WhatsApp workflow for clinic inquiries that turns approved questions and intake into a useful staff handoff.

Approved clinic questions and structured intake ready for staff

01

Project overview

Clinic inquiries organized for the person who handles the next step

Services: AI Receptionists ; AI Workflow Automation ; AI ERP & CRM
Delivery: WhatsApp inquiry, intake, and staff-handoff workflow. Portfolio alias.

02

The Problem

Clinic WhatsApp conversations mix several requests: service information, appointment preferences, intake details, and questions that require staff attention.

If those messages remain an inbox, the receptionist must read the thread, recover the contact details, identify the request, and copy the context into the next workflow. Repeated questions also compete with messages that need a person.

03

What We Delivered

CareLine answers approved non-clinical questions, collects contact and intake information, and prepares the next step for staff.

The workflow covers:

  • Service, location, availability, and preparation questions.
  • Patient contact details and the reason for inquiry.
  • Preferred appointment or callback times.
  • Routing of sensitive, uncertain, or out-of-policy messages.
  • A conversation summary, captured fields, and a handoff reason.

A booking request and a confirmed appointment are kept distinct. The assistant should not confirm a live slot before the scheduling workflow accepts it.

04

Messaging Implementation

The messaging foundation uses Meta WhatsApp webhooks, incoming text handling, HMAC signature checking, and outbound replies through the Graph API.

OpenAI generates short channel-appropriate replies, with a fallback path when the model service is unavailable. Transcript entries preserve incoming and assistant messages for the application view.

The demonstration covers the conversation and handoff. Scheduling acceptance and durable CRM records are separate integrations.

05

Engineering Decisions

Capture the request before handing it over. The next staff member needs the reason for contact, relevant details, and the requested next action.

Keep clinical decisions with staff. The workflow does not diagnose, prescribe, or replace emergency care.

Verify channel events. Webhook authenticity checks are separate from AI reply generation.

Keep the confirmation boundary explicit. Collecting a preferred time does not mean the slot is booked.

06

Results

The delivered workflow turns the inquiry into structured intake and a defined next action for staff. It also retains the conversation context needed to continue the exchange.

No response-time, booking-conversion, or staffing-cost figure has been supplied. Those measurements should distinguish automated replies, staff handoffs, booking requests, and confirmed bookings.

07

Technology stack

Meta WhatsApp Graph API; Next.js, React, TypeScript; OpenAI reply generation; webhook signature validation and transcript handling.

08

Example value calculation

Hypothetical planning example, not a measured project result or a performance forecast. All volumes and timings below are assumed.

Capture contact details and the request, then review the summary and next step.

  • Workload: 300 eligible non-clinical inquiries per month.
  • Manual handling: 5 minutes per item.
  • Assisted handling, including human review: 2 minutes per item.
  • Additional exception handling: 90 minutes in the same period.
  • Modeled difference: 300 x (5 - 2) - 90 = 810 minutes (13 hours 30 minutes) per month.

The example reserves another 90 minutes per month for incomplete intake and routing corrections. Clinical conversations and appointment fulfillment are excluded. This is potential staff capacity, not cash savings; setup, training, and software costs are excluded.

How to validate: Measure staff handling minutes for eligible administrative inquiries, including summary review and rework. Track handoff completeness, misrouting, booking requests, and confirmed appointments separately.

09

Project at a glance

WhatsApp intake with the context ready for staff.
CareLine handles approved clinic questions, captures inquiry details, and prepares booking or callback requests and staff handoffs. Clinical decisions remain with the clinic.

10

Project summary

CareLine handles the first part of a clinic WhatsApp inquiry: approved questions, contact details, the reason for contact, and the requested next step. Staff receive the context rather than an unprocessed thread.

Build a system around your workflow.

Tell us where the work gets stuck. We will map the first release, integrations, cost, and timeline.

Discuss a similar project