What gets reviewed
- Your last quarter of incidents and pages: what fired, how long it took to understand, who had to be woken up
- Your runbooks and alerts: which are current, which are stale, which don't exist
- Your tooling: how on-call engineers get from an alert to the right data today
- Interviews with two or three people on the rotation
What you get
Report
A ranked automation plan
- Your recurring incident classes, ranked by frequency, time to diagnose and impact
- For each: what can be computed deterministically, what needs judgment, and what it would take
- Quick wins your team can do without HASHJOIN
Prototype
One working analyzer
- A parser or analyzer for the top-ranked incident class, run against your own data
- Yours to keep, whether or not you go further
How it works
Kickoff day 1
Agree on access (read-only is enough), the incident window and who to interview.
Review days 2–8
Go through incidents, runbooks and tooling; run the interviews; build the prototype.
Readout days 9–10
A one-hour session walking through the report and the prototype, then a written copy.
Price
$7,500, fixed, paid at kickoff.
If you go on to an ops automation build within 60 days, the full $7,500 is credited toward it.