Version Control & Collboration
There is one Production model (the only version visible to business users) and a single shared Staging model (where all Data Experts collaborate).
There are no personal branches: every Data Expert sees everyone else’s in‑progress changes in Staging. When a change set is ready, it’s published (promoted) to Production.
Collaboration Practices
- One shared Staging = communicate early.
- Announce large or impactful changes in the team channel/ticket.
- Small, incremental publishes is recommended over big-bang releases.
- Test well before publishing.
Environments
Production
Audience: Business users (read-only), Data Experts (read + historical diffs).
State: Always stable and verified.
Goal: Trustworthy, org-wide analytics.
Staging (shared)
Audience: Data Experts only.
State: Mutable, continuously evolving.
Goal: Fast iteration, testing, and collaboration before publish.
Updated 5 months ago
