Onboarding runbook · if you hire me
My first 90 days on your team
A draft. By the end of week one I would rewrite it around what your team needs.
Docs · the plan
01Days 1–30
Listen and map
Understand before changing anything.
- Get access the proper way: my own accounts and keys, least privilege from day one.
- Inventory every service: where it runs, how it deploys, who owns it, what it costs.
- Read the incident history and shadow on-call.
- Ship one small, safe pipeline improvement to learn the release path end to end.
02Days 31–60
Make releases predictable
Fewer surprises on deploy day.
- Turn the deploy process into one template with approvals, health checks and rollback.
- Close monitoring gaps: alert on what users feel, not every CPU spike.
- Run a restore from backup, and write down how long it really takes.
03Days 61–90
Make it measurable
Numbers the team and management can both read.
- Report deploy frequency, change failure rate and time to restore each month.
- Codify at least one server build (Ansible or Terraform), so a rebuild is not tribal knowledge.
- Write runbooks for the five most common alerts and hand them to the team.
Docs · what I won't do
- Rewrite everything in the first month.
- Change production without a rollback path.
- Add a tool before there is a problem it solves.
Want to adapt this to your team?