Docs
Guides

Guides

Practical workflows for instrumenting a product and operating Customer Success as Code.

These guides take you from product facts to production operations. Complete the 5-minute quickstart first so you have an authenticated CLI and a generated Kite project.

Choose a guide

GoalGuide
Choose customer IDs and product eventsIdentity and event taxonomy
Define activation, risk, and expansionActivation, churn, and expansion
Build an explainable scoreHealth score
Import history and rebuild stateBackfill and recomputation
Promote configuration safelyCI/CD and environment delivery
Consume signed outbound eventsWebhooks and signatures
  1. Choose one stable account or workspace ID.
  2. Instrument the smallest set of events that proves one customer outcome.
  3. Model that outcome and validate locally.
  4. Deploy to test, send representative events, and inspect the resulting state.
  5. Add health, automation, and production delivery only after the model is explainable.
kite validate --strict
kite dev
kite deploy --env test
kite customers inspect customer_123 --env test

Use the Config Reference for DSL fields, the CLI Reference for command options, and the SDK Reference for backend delivery behavior.

Start from observed behavior

Model one measurable outcome using events your product already produces reliably. A small, tested model is more useful than a broad taxonomy based on assumptions.

On this page