Build Trust with Your Users Workshop
summary
Workshop lays out concrete reliability practice: define SLIs and SLOs (examples like 99.9 targets), convert targets to an error budget and use 28-day rolling windows with short 1–12 hour windows for alerting, measure availability and latency from the browser, and make release velocity depend on remaining budget. It demos a shipped product, Polaris SDK 7KB gzipped, that captures Core Web Vitals, sends real-time SLO measurements, integrates with PagerDuty/Jira/Datadog/Slack, and provides AI recommendations for metrics and objectives.
AI-written summary · about
chapters (11)
- 0:00Welcome
- 2:26Web apps need SLOs
- 6:18Features versus reliability
- 11:11Reliability targets
- 17:55Error budgets
- 22:51Choosing SLIs for happiness
- 27:21Aspirational SLOs
- 33:00User journeys with SLIs
- 38:38Polaris to escape hell
- 42:27Polaris for real user performance
- 47:21Q&A
