Live signal warming up.
90-day uptime per capability, plus every incident we've opened in the last quarter. The signal here comes from our internal synthetic-check workflow, one ping per vendor per minute against the real APIs we depend on.
Voice line
Inbound calls, AI answering, transfers and messaging.
Inbound replies and outbound sends from Workhand.
SMS
Inbound and outbound SMS over Twilio.
Calendar sync
Google Calendar bookings, reschedules, conflict checks.
AI
Anthropic primary, OpenAI fallback. Drives every reply and call.
Customer dashboard
/client/* pages, sign-in, settings.
Google reviews
Inbound reviews + drafted responses.
We post an entry to the changelog on any incident that affects customer-visible behaviour.
Affected customers get a direct email within 30 minutes of acknowledgement. P0 incidents trigger an SMS alongside.
If you need a real-time check during an incident, email felix@workhand.co.uk , first acknowledgement within 5 minutes during work hours.
One email per incident at P1 or higher, nothing else. Unsubscribe by replying or emailing felix@workhand.co.uk.
Every incident we opened in the last 90 days, source = audit_log. We don't edit these after the fact.
No incidents in the last 90 days. The receipts grow here as they happen, we don't edit them after the fact.