Skip to main content

discovery

🔍 Discovery Issue​

This issue kicks off a lightweight discovery cycle. Use it to define why we are doing discovery and what we expect to learn. All detailed findings go into the Discovery Sheet, not this issue.

Owner Olivia

Olivia — JTBD‑2: Detect and Diagnose Issues Quickly

Why are we doing this?​

Owner Olivia spends valuable time during incidents jumping between dashboards, logs, and traces to understand what is happening. We lack clarity on whether a correlated view, alerts, or another mechanism best helps her diagnose issues quickly and confidently.

What do we want to learn or validate?​

  • Does Olivia prefer proactive alerts, a guided incident view, or raw dashboards during incidents?
  • What signals matter most in the first 5 minutes of an incident?
  • What level of correlation (logs, metrics, traces) is required to build trust?
  • Can we reduce tool switching during triage?

Planned discovery activities​

  • Interview Owner Olivia about incident response workflows
  • Review recent incident timelines and decision points
  • Run concierge tests with manually prepared incident snapshots
  • Validate feasibility of correlating signals with existing data

Expected discovery deliverable​

  • Discovery Sheet summarizing insights, evidence, and decisions
  • Clear recommendation: build, pivot, or continue learning

Definition of Done (Discovery)​

  • Discovery activities completed
  • Discovery Sheet created in /docs/discovery/...
  • Decision documented and shared
  • Link to Discovery Sheet added here
  • Feature issue created (if appropriate)

💡 Note:
Keep this issue lightweight.
All detailed evidence and reasoning belong in the Discovery Sheet.