Observability and On-call Setup
Observability and On-call Setup gives your team monitoring, logs, alerts and a working on-call process. We build dashboards and alerting tied to service level objectives, write runbooks for the alerts that matter, and set up routing so the right person is paged for the right problem.
- Typical duration
- Typically 3 to 5 weeks, depending on platform size, confirmed in the written scope
- Format
- Fixed scope, delivered in your tooling and repositories, with training
- Led by
- A named senior engineer
Who this is for
- Teams that find out about outages from customers
- Teams drowning in noisy or ignored alerts
- Companies that need a first on-call rota that people can live with
- Teams preparing for enterprise customers who ask about uptime and response
What is included
Metrics and dashboards
Prometheus and Grafana, or your preferred managed tool, with dashboards engineers actually use.
Logging and tracing
Centralised logs and, where it pays off, distributed tracing with OpenTelemetry.
Service level objectives
SLIs and SLOs for your critical services, agreed with the business, with error-budget visibility.
Alert design
Symptom-based alerts with severity levels, replacing noisy cause-based alerts.
Alert routing and paging
Routing to chat and a paging tool by severity, with escalation and quiet-hours rules.
Runbooks
A runbook for every paging alert, linked from the alert itself.
On-call process
Rota design, handover, incident roles and a simple post-incident review template.
Training
A session for your engineers on operating the stack and responding to incidents.
What you receive
- ✓A working monitoring, logging and alerting stack
- ✓Dashboards and SLOs for your critical services
- ✓Runbooks linked from alerts
- ✓A documented on-call and incident process
- ✓A handover and training session
What we need from you
- •Read access to the platform and existing monitoring
- •A named service owner for each critical service
- •A decision on paging tool and chat channels
- •Time from engineers for design and tuning sessions
Not included
- –Penetration testing or formal security audits
- –Compliance certification (we implement technical controls, we do not issue certificates)
- –Providing 24/7 on-call staff (see the Managed DevOps plans)
How it runs
- 01
Discover
We review current monitoring and incidents, identify critical services and agree the scope in writing.
- 02
Design
We define SLIs, SLOs, alert policy and routing with your engineers.
- 03
Build
We implement the stack and alerts as code in your repositories.
- 04
Tune
We run the alerts against real traffic and remove noise before anyone is paged.
- 05
Hand over
We deliver runbooks and training, and start the on-call rota with you.
Every package starts with a short scoping call. We then confirm scope, deliverables and timeline in writing, and you decide whether to go ahead. Prices are quoted for that fixed scope, not as a rate card.
Who you will work with
Engagements are led by a named senior engineer, not passed to an anonymous team.
CEO, DevOps / SRE / Cloud Consultant Engineer
You keep everything
No lock-in, by design. If you stop working with us, nothing you rely on stops working.
Your accounts
Everything runs in cloud accounts and subscriptions you own and control.
Your code
Infrastructure code, pipelines and configuration live in your repositories.
Your documentation
Architecture notes and runbooks are written for your team and handed over.
A defined exit
We agree the handover steps in writing, and access is removed when the work ends.
Related services and technologies
Frequently asked questions
Do you only use open-source tools?
Can you reduce alert noise on an existing setup?
Do you provide on-call engineers?
What are service level objectives for?
How will we know it works?
Get a written scope for Observability and On-call Setup
Book a short scoping call. You receive scope, deliverables and timeline in writing, and you decide whether to go ahead.