215 words. Estimated reading time: 1 min.
Data
Data architecture overview
Data used by Onboarding Specialists comes from three main sources:
Billing Postgres <PrivateLink url="https://billing.posthog.com/admin/"><em>admin panel view</em></PrivateLink>
- Customer account, subscription, and invoice type data
- Usage reports and consumption metrics
- Revenue amortization calculations
- Billing forecast / spike calculations
Production Postgres _admin panel view (US)_
- Organizations and projects configuration
- User accounts and permissions
- Product settings and feature flags
- Warehouse tables, pipeline source/destination info
ClickHouse
- Event and person data for all teams (projects)
- person to distinct_id mappings
Query capabilities
- Metabase queries production databases directly but cannot combine Postgres and ClickHouse in a single query
- PostHog analytics limited to Team 2 data, but can query across databases in a single query
- Cross-organization analysis requires Metabase for customer event analysis, including:
- Library usage breakdowns
- Event volume metrics
- Implementation diagnostics
Key data tables
For onboarding analysis, these tables provide essential data:
| Table | Purpose | Key fields | |-------|---------|------------| | invoice_with_annual | Billing data with revenue amortization | Revenue (mrr), billing period, type (annual, completed, upcoming, etc) | | posthog_organization | Org-level configurations | Settings, feature access, creation date | | posthog_project | Project/team settings | Project configuration, team members | | billing_spike | Usage anomaly detection | Spike timestamps, magnitude, affected metrics |