Metrics and Dashboards
Notebook labs for metric forensics
Use notebooks responsibly to explore spikes without turning ad-hoc queries into permanent dashboards by accident.
- Format
- Self-paced
- Duration
- 3 weeks, async-first
- Skill
- Intermediate
- Stack
- Python, Prometheus HTTP API
Tuition (informational): KRW 240,000
Request a syllabus conversation
Notebooks tempt copy-paste chaos. You learn patterns for ephemeral exploration: parameter hygiene, saving conclusions instead of queries, and linking back to canonical dashboards. Labs use Python against Prometheus APIs with gentle rate limits so habits respect shared clusters.
What is included
- Notebook templates with guardrail cells
- Exercises on parameterizing variables teams can reuse
- Ethical scraping reminders and backoff snippets
- Short module on turning insights into tickets, not mystery files
- Pairing on reviewing a teammate's exploratory notebook
- Visualization choices that survive grayscale printing
- Wrap-up on archiving notebooks without shame
Outcomes
- You finish a spike notebook others can replay without your shoulder.
- You adopt a naming scheme for temporary cells versus canonical work.
- You identify one anti-pattern your team notebook gallery currently rewards.
Instructor of record
Content strategist who once deleted a hundred orphan notebooks and lived to tell the tale gently.
Yuri Tan
Primary feedback on labs
Participant questions
If you can open a CSV with pandas once, you are fine. We do not teach data science theory.
Recent voices
“Guardrail cells saved our shared cluster from another ad-hoc `topk` storm during release week.”