I diagnose scenarios that break, and I monitor the ones that keep succeeding while the result stops being correct.
Diagnosis: you send a blueprint export plus execution history, I return a written report within 48-72h of the package being complete. 25-point checklist: scheduling and overlap, error routes and rollback, retry idempotency, silent quality drift, operations cost traps, observability. Findings ranked by severity, top 3 fixes, 30-day plan. Fixed price, async, no calls, I never touch your live scenario. One autopsy covers one process on one platform, up to 3 connected workflows or 40 steps, on 7 days of run history.
Monthly care: content-level canary checks, dead-man switch on schedules, alerts, monthly written health report. For teams running client-facing scenarios that nobody re-checks after launch. Care covers up to 3 loops. It is monitoring, not repair: you get the alerts, the monthly written health check and the setup instructions for the canaries, which you connect yourself. No changes on your production, no 24/7 response, no SLA.
The gap I cover, in one sentence: monitoring tells you the run finished, I check whether what it produced is still correct.
DM me here with a screenshot or description and I will start with a free 3-point mini-read, capped at 20 minutes, so you can judge quality before any money moves.
This is my one listing in this category. I update it in place rather than reposting.