Continuity After the Only DevOps Left
Client: Bootstrapped B2B SaaS (anonymized)
Package: Managed Support
Timeline: ongoing (started after a short onboarding)
Outcome: Cluster upgrades and incident help covered without an emergency full-time SRE hire
Problem
The only engineer who understood production Kubernetes left. Product still shipped; cluster drift and deferred upgrades became the quiet risk. Hiring a senior SRE would take months and more cash than the runway liked.
Approach
- Short onboarding: access model, blast radius, upgrade debt, alert noise
- Monthly hour bucket for changes, reviews, and advice
- Planned upgrade path instead of “hope the node group survives”
- Async channel for operational questions inside the SLA window
Results
- Two minor version upgrades completed without a customer-facing outage window surprise
- Incident response had a named escalation path instead of “whoever is awake”
- Founder avoided a panic hire; support hours stayed below a full-time salary
Takeaway
Managed Support is for ownership and continuity — not a substitute for product engineering. Scope stays platform-adjacent so the retainer does not turn into an unbounded outsourcing contract.
Talk about Managed Support
Tell us hours, blast radius, and whether you need a Health Check first.