# Observability & SRE

URL: https://avispa.ai/agent-teams/observability-sre/

Platform & Engineering team

Runs observability on the Grafana LGTM stack on OpenShift: instrumentation, pipelines, dashboards, SLOs and alert quality. It works with your SRE team, or runs the practice if you have none.

## Who's on the team

- **Observability Lead** · Lead · Leads the practice and owns every approval gate.
- **Grafana Dashboards Engineer** · Builds and maintains dashboards and alerting in Grafana.
- **Loki Logs Engineer** · Works on logs in Loki and the pipelines that feed them.
- **Mimir Metrics Engineer** · Runs metrics in Mimir and their cost.
- **Observability Platform Engineer** · Watches the OpenShift platform and posts a daily cluster health digest.
- **OpenTelemetry Engineer** · Instruments services with OpenTelemetry.
- **Reliability Engineer** · Defines SLOs and reviews SLO burn and alert noise every week.
- **Tempo Traces Engineer** · Works on distributed traces in Tempo.

## How a team arrives

1. 1 · **Import** · Arrives switched off
2. 2 · **Connect** · Credentials stay with you
3. 3 · **Discovery** · Learns how you work
4. 4 · **You approve** · Its plan and its limits
5. 5 · **Activate** · Routines switch on

## Controls it works inside

- Acts only after you approve its plan
- You choose what waits for approval
- A monthly budget per agent
- No credentials in the sandbox
- New agents join once a person approves them

- Only the Grafana Dashboards Engineer, the Mimir Metrics Engineer, the Reliability Engineer and the Observability Lead can make changes. Every other grant is read-only.
- No alert, dashboard, pipeline or paging change before you approve the operating plan.
- Every page maps to a human action, and alert noise is reviewed as a budget.

**More: when and where it works, and what it documents**

## When it works

- **Daily Cluster Health Digest** · Weekdays at 08:00
- **Monthly Cardinality and Cost Review** · The 1st of each month at 09:00
- **Weekly Alert Noise Review** · Fridays at 15:00
- **Weekly OpenTelemetry Coverage Report** · Mondays at 09:00
- **Weekly SLO Burn Review** · Mondays at 10:00

Routines arrive switched off. You turn each one on.

## Where it works

- Jira
- Cribl Stream
- Grafana
- Loki
- Mimir
- Tempo
- OpenShift
- Slack

Each connection is granted per agent, read-only or full. [All integrations](https://avispa.ai/integrations/)

## What it documents

- A daily cluster health digest.
- Weekly SLO burn, alert noise and OpenTelemetry coverage reports.
- A monthly cardinality and cost review.

[All agent teams](https://avispa.ai/agent-teams/) · [Next: Operations & Vendor Management](https://avispa.ai/agent-teams/operations/)

## Onboard Observability & SRE with us.

[Book a demo](https://avispa.ai/book-a-demo/) · [Watch it work](https://avispa.ai/#watch)