SRE The production that goes down and no one knows why.
When something breaks in production, you find out the hard way: a customer complains, and only then does the team scramble. There is no observability to show the problem before the user feels it, no SLO to say what “good enough” means, and every incident becomes firefighting, with no learning to prevent the next one.
What the engineer delivers.
- Real observability: metrics, logs and traces that point to the cause, not just the symptom.
- SLOs and SLIs defined with you, tied to what the business feels, not just CPU and uptime.
- Incident response with clear roles: who acts, who communicates, who decides.
- Postmortems that end in an applied fix, with an owner and a deadline.
Bring the SRE request as it stands.
Describe the problem the way it sits in your head. Triage turns it into a plan, and the person reading on the other side is an architect.