The seven dependencies and what fails
Seven dependencies, sorted by damage: Microsoft Azure (Canada Central) is the one that can stop the clinic and Microsoft's status page is the source of truth; Stripe and Daily.co each stop one feature (card payments, video visits); Alberta H-Link, Anthropic, Twilio and SendGrid degrade — claims queue and go when H-Link returns, the AI stops answering while booking by phone and at the desk carries on, unsent texts or emails are visible rather than silent and the other channel keeps working.
2705
If this fails, the clinic stops
Microsoft Azure, Canada Central — every record, appointment, claim and the application itself. If it goes down you cannot reach Qlynic at all; nothing is lost, the data is where it was, and Microsoft's status page is the source of truth, not ours.
If this fails, one thing stops
- Stripe — cards cannot be charged. Cash and AHCIP billing are unaffected, and the invoice is still raised.
- Daily.co — video visits cannot start. In-person clinic runs normally, and the appointment can become a phone visit without losing the booking.
If this fails, it degrades
- Alberta H-Link — claims queue locally and go when it returns; nothing dropped, nothing re-entered; charting and booking untouched.
- Anthropic — the AI booking agent, the scribe's drafting and the advisor stop answering. Booking by phone and at the desk carries on, because the schedule was never the AI's to hold.
- Twilio — texts do not send; email reminders still do, and the unsent list is visible rather than silent.
- SendGrid — email does not send; SMS still does, and again the unsent list is visible.
Six of the seven cannot take the clinic off the air. A live system status page is marked Soon in the product menu, and Qlynic does not publish an uptime figure.
Where to look