The message nobody saw
A parent whose child spiked a fever messages the clinic on WhatsApp. No reply until morning. By 10 PM they've booked the clinic that answered.
Salma replies in seconds on WhatsApp, Instagram and Messenger, and books the appointment before they try another clinic — day and night, in their language.
Three parents, three apps, the same minute — all booked. Tap the tabs to watch each conversation.
Every unanswered message is a patient who books with whoever replies first.
A parent whose child spiked a fever messages the clinic on WhatsApp. No reply until morning. By 10 PM they've booked the clinic that answered.
The front desk opens to a stack of unread messages and answers the same four questions — price, time, location, insurance — one by one, all day.
A no-show nobody chased, a cancelled slot nobody re-filled, a follow-up nobody sent. Each one is paid staff time and an empty chair.
No apps to learn and no scripts to write. Salma works on the channels your patients already use.
On WhatsApp, Instagram or Messenger — any hour, in Arabic, English, Hindi or Russian. Salma replies in seconds.
She answers their questions, checks your live calendar, and books the right slot with the right department — no double-bookings.
The doctor gets a short summary before the visit, your staff sees every chat, and anything urgent is flagged to a person straight away.
Salma reads your bookings, no-shows and revenue, then writes you the moves — who to hire, which service to grow, which to close. Decisions, not dashboards.
The engineering that makes Salma safe in front of patients.
A built-in safeguard makes every booking run exactly once. A crash or a re-sent message can never double-book or double-send.
A deterministic emergency screen runs before the model replies. Ambiguous cases escalate to humans with severity flags.
Structured key facts per patient, recalled across channels — the patient who messaged on WhatsApp and later writes on Instagram is recognised.
WhatsApp, Instagram and Messenger — one patient, one unified conversation, one inbox for your staff.
Each clinic's credentials, conversations and data are strictly separated. No shared accounts, no global fallback.
Every action logged, credentials encrypted at rest, data exported or deleted on request.
Arabic and English side by side, patients from everywhere, WhatsApp as the front door — what Salma was built around.
Full right-to-left Arabic, plus English, Hindi and Russian, auto-detected per message — matching who actually walks into UAE clinics.
Data handling designed around Federal Decree-Law 45/2021 — purpose-bound processing, clinic-controlled data, deletion honoured on request.
Built on Meta's WhatsApp Business Platform with Embedded Signup — your own number and account, connected in minutes, tokens never exposed.
Reminders and recalls respect evenings, weekends and local rhythms — proactive without ever feeling like spam.
Multi-tenant from day one — groups and chains run every branch under one roof, with each clinic's data strictly isolated.
We set up your first clinic with you — from connecting your number to the first booked patient, in days, not months.
The name comes from the Arabic root salima — to be safe and sound. That's the promise we build to: your patients, your calendar and your data, in safe hands.
Salma is an AI healthcare operation manager — it runs a clinic end to end. It answers patients, triages symptoms and books appointments 24/7; confirms with the doctor and summarises the case; and runs the business — telling you who to hire, who to let go, and which services to grow, close or extend.
Yes. Salma is an operation manager, not just a receptionist. It studies your booking, no-show and revenue data and recommends business moves: who to hire, who to release, which services to expand or sunset, and which campaigns and strategy to run.
Salma triages the patient, confirms the appointment with the right provider, and summarises the symptoms and schedule for the doctor — so the clinician walks in already briefed.
WhatsApp (through Meta's official Cloud API), Instagram DMs and Facebook Messenger. One patient conversation stays unified across all of them.
Every booking runs through a safeguard that lets it happen exactly once. A re-sent message, a crash, or two requests at the same time never create a second booking or a duplicate message.
Salma runs a deterministic red-flag classifier and a focused triage step before the model replies. Ambiguous emergencies are escalated to staff with severity flags — Salma does not freewheel on potential emergencies.
Yes. Each clinic's credentials, conversations and data are strictly isolated per-tenant. There is no shared account or global fallback.
Yes. Salma stores structured key facts per patient and surfaces cross-channel recall — so a patient who messaged on WhatsApp and later writes on Instagram is recognised with their history intact.
English, Arabic (full right-to-left), Hindi and Russian, auto-detected per message.
Through Meta's WhatsApp Embedded Signup inside the Salma dashboard: you authorise Salma to manage your own WhatsApp Business Account, pick your number, and Salma registers it and subscribes the webhook — your token never reaches the browser.
Book a demo — we onboard your first clinic with you: create your clinic, connect your channels, and Salma starts running.
Book a demo — we'll connect your WhatsApp number, onboard your first clinic with you, and let Salma take the night shift.