An integration is not finished when it goes live — it is finished when it keeps working month after month without anyone thinking about it. We monitor live integrations, respond to incidents against agreed targets, and manage the slow drift of platform APIs so a connection that worked in January still works in December.
What we look after
- Integrations we built. Covered from go-live under a support agreement, with the monitoring and alerting already in place.
- Integrations you inherited. Built by a previous agency, a contractor or an in-house developer who has since moved on — often with little documentation. We audit it, wrap it in monitoring and take it under an agreement.
- A mix. Most clients have several connections of different ages and origins. We can hold the whole set rather than just the ones we wrote.
Monitoring and alerting
- Health checks on every data flow, not just a ping — we verify that records are actually moving.
- Thresholds on error rate, latency and queue backlog, tuned to what is normal for your traffic.
- Alerts to the channel you already watch — email, Slack or Teams — the moment something moves outside those thresholds.
- A status view of every integration and a weekly digest, so a slow degradation is visible before it becomes an incident.
Incident response
When a transfer fails we triage it, apply a fix or a safe rollback, replay any records that were missed, and send a short written explanation of what happened and what stops it recurring. Response and fix targets depend on the tier you choose:
| Standard | Priority | Critical | |
|---|---|---|---|
| Cover hours | Business hours, Mon–Fri | Extended hours, Mon–Sat | 24 / 7 |
| Monitoring & alerting | Health checks, daily digest | Real-time, on-call alerting | Real-time, on-call alerting |
| Response — integration down | Next business day | 4 business hours | 1 hour |
| Response — degraded / partial | 2 business days | 1 business day | 4 hours |
| Response — question or minor issue | 3 business days | 2 business days | 1 business day |
| Target fix or workaround — integration down | 3 business days | 1 business day | 8 hours |
| Included change work | Up to 2 hours / month | Up to 1 day / month | Up to 2 days / month |
| Platform API version migrations | Quoted per change | Included | Included |
| Quarterly review | — | Yes | Yes |
Response time is measured from when an issue is raised through the agreed channel. Tiers and thresholds are set per engagement — the table above is our usual starting point.
Keeping up with change
- Platform APIs are versioned and change on their own schedule. Integrations under an agreement are reviewed each release cycle and migrated ahead of deprecation dates, not after something breaks.
- When you patch or upgrade one of your own systems, we run the integration against it in staging first and confirm nothing has moved.
Taking over an existing integration
We start with a fixed-price audit: how it actually works, where the single points of failure are, what is undocumented, and what it would take to bring it up to standard. You get a written report and a quote. If you go ahead, we add monitoring, write the runbook and put it under a support agreement — no rebuild required unless the audit finds one is genuinely needed.
Talk to us about support
Tell us what you are running and how it is currently watched. We will propose a tier and, where needed, a short audit first. Get in touch or call 01303 883111.