Lovable
Stream every call, lead and booking from your AI receptionist into an app you build in Lovable: a live dashboard, a lead board or a client portal, in your own design.
lovable.dev ↗Apps built in Lovable have a backend that can receive HTTP requests, and Cyberstaff’s Custom Webhook integration sends every call outcome as a JSON POST to any URL you give it. Point one at the other and your Lovable app receives each event seconds after it happens, with automatic retries and a delivery log on the Cyberstaff side. No marketplace app, no API polling, nothing to install in Lovable.
1. Before you start
- A Cyberstaff account with an active AI receptionist number.
- A Lovable project with its built-in backend enabled, so your app can expose an API endpoint that receives HTTP requests. Any app you build in Lovable can do this: just ask Lovable to add an endpoint (see the prompt below).
2. Ask Lovable to build the receiving end
Paste this prompt into Lovable (edit the dashboard part to taste), and it will build the endpoint, the storage and the pages around it:
Add a public API endpoint at /api/cyberstaff-webhook that accepts
POST requests with a JSON body shaped like:
{ "event": "call.completed", "schema_version": "1.1",
"timestamp": "...", "data": { ... } }
Store every event in a table, then build a dashboard page that lists
calls using fields from "data" (all snake_case): customer_name,
customer_phone, call_timestamp, duration_seconds, call_reason,
call_summary, transcript_text, lead_score and lead_rating
(hot / warm / cold), plus call_url linking back to the call.
Other event types ("booking.created", "booking.confirmed",
"booking.declined", "message.conversation.completed") arrive at the
same endpoint; store them too and show bookings on their own page. 3. Connect Cyberstaff to your Lovable app
- Ask Lovable to add the receiving endpoint to your app using the prompt above, then publish your app and copy the endpoint URL.
- Log in to your Cyberstaff dashboard, open the Integrations page and connect the Custom Webhook integration.
- Paste your Lovable endpoint URL, toggle on the events you want, and save.
- Optionally enable request signing (HMAC) on the connection and ask Lovable to verify the signature before accepting a request.
4. Choose which events arrive
Each event is an independent toggle on the connection, and every one arrives in the same envelope: { event, schema_version, timestamp, data } with flat snake_case fields in data, ready to store and render:
| Event | What your app receives |
|---|---|
| Call completed | Fires the moment a call ends, with the AI summary, full transcript, lead score and hot/warm/cold rating, caller details, call duration, forwarding and transfer outcomes, and a link back to the call in your dashboard. |
| Booking requested | A caller asked to book: arrives with the service, the availability windows the caller gave and the next available slot. |
| Booking confirmed | A booking was confirmed, with the selected time and duration. |
| Booking declined | A booking request was declined, with the reason. |
| Messaging conversation completed | An SMS or WhatsApp thread wrapped up: the channel, customer details, message count and an AI summary of the thread. |
Use the Send test button next to each enabled event to push a sample payload at your Lovable endpoint and check it lands before any real call comes in.
5. What people build with it
Every call your AI receptionist answers appears in your Lovable app seconds after it ends, with the summary, lead rating and transcript, no copy-paste and no CSV exports.
Filter on lead_rating to surface hot leads, and use callback_phone and customer_name to build a call-back queue your team works through each morning.
Building an app for someone else? Stream their receptionist activity into the portal you host for them, styled exactly how you want it.
Webhook not arriving in your Lovable app, or unsure how to shape the dashboard? Our team can help debug the delivery or check your event toggles.