Week 1 · Signal inventory
Catalog latency, traffic, errors, and saturation for one bounded service — including what you are not measuring yet.
Ops Signal Studio
Eight weeks of application analytics for microcloud operations health — from first signal map to a review ritual your team will actually keep.
What you will practice
- Build a service-level signal map for one production application
- Separate cosmetic green from operationally meaningful green
- Write baselines that account for Korea business-hour traffic and weekend quiet
- Facilitate a thirty-minute weekly health review without status theater
- Document limitations so on-call successors inherit honesty, not folklore
Modules
Each week pairs a short live studio with async labs you can run against a sandbox or your own staging stack.
Week 2 · Baseline honesty
Construct baselines that survive deploys, cache warmers, and promotional traffic without rewriting history.
Week 3 · Sampling & shape
See how aggregation windows flatten incidents and when percentile choices flatter the dashboard.
Week 4 · Error narratives
Turn error rates into partner-ready stories with clear ownership and recovery hints.
Week 5 · Saturation sketches
Draw queue and connection pressure before customers feel pauses.
Week 6 · Review rituals
Design a weekly ops health review that fits a microcloud team of three to twelve people.
Week 7 · Limitation notes
Publish the known blind spots: cold starts, multi-tenant noise, and third-party dependencies.
Week 8 · Studio showcase
Present your signal map and receive structured critique from instructors and peers.
Instructor
Soo-yeon Park spent a decade on reliability teams for regional SaaS products before founding the Ops Signal Studio curriculum. She teaches from Gumi with guest critiques from Seoul-based platform leads.
Informational pricing
No checkout on this site — figures help you plan. Enrollment happens through our team.
Individual seat
₩890,000 / cohort
- Live studios + recordings
- Async lab feedback
- Private community channel
Team pod (up to 5)
₩3,600,000 / cohort
- Shared signal-map critique
- One private office hour
- Template pack for your stack
FAQ
Do I need Prometheus or a specific vendor?
No. Labs accept exports from common stacks. We focus on interpretation habits that transfer when tools change.
Is this suitable for brand-new engineers?
You should already ship or support at least one production service. Absolute beginners may feel lost in week three when we discuss sampling bias.
What is a real limitation of this course?
We do not deep-dive multi-region failover analytics or Kubernetes autoscaler internals. Those deserve dedicated workshops; treating them lightly would create false confidence.
How are sessions timed for Korea?
Live studios run early evening KST. Recordings post within 24 hours for teammates abroad.
Learner notes
“Week 6’s review ritual replaced our noisy Slack dump. The template still sits at the top of our ops channel.”
★★★★☆
Solid on baselines. The showcase feedback was blunt in a useful way; I rewrote two panels before presenting to leadership.