Skip to main content

​
Dashboard Endpoints → BigQuery Data Sources

​
Summary (3–6 bullets)

  • Maps key dashboard endpoints to their underlying BigQuery tables/views.
  • Clarifies that dashboards read downstream processed/analytics objects, not intake raw bridge tables.
  • Highlights source boundaries for CEO metrics, top entities, growth/loss, and commissions.
  • Provides a quick drift-risk perspective for dashboard data dependencies.

​
When to use this (3–6 bullets)

  • When debugging dashboard metric mismatches or stale values.
  • Before changing query functions or underlying BigQuery views/tables.
  • During onboarding to understand endpoint-to-data lineage quickly.
  • When validating whether intake pipeline changes can affect dashboards directly.

​
What you’ll walk away with (2–5 bullets)

  • A concise source map for major dashboard APIs.
  • Confidence on where dashboard data is and is not sourced from.
  • A fast reference for impact analysis before backend/data changes.

​
1. CEO Metrics (/api/v1/ceo-metrics)

Function: get_ceo_metrics() / get_ceo_metrics_range() / get_ceo_snapshot_from_view() Source Tables/Views:
  • payroll-bi-gauntlet.payroll_analytics.ceo_snapshot (view) - Financial metrics (gross_payout, chargebacks, net_payout)
  • payroll-bi-gauntlet.payroll_raw.stage1_snapshots (table) - Distinct counts (total_businesses_ytd, total_employees_ytd)
  • payroll-bi-gauntlet.payroll_processed.stage3_snapshots (table) - Agent commissions for owner commission calculation
Uses stage1_bridge_rows? No - reads from stage1_snapshots (downstream view) and ceo_snapshot (analytics view) Uses transaction_events_raw? No - reads from processed/analytics views

​
2. Agent History (/api/v1/agent-history/{agent_id})

Function: get_agent_history() Source Tables/Views:
  • payroll-bi-gauntlet.payroll_processed.stage3_snapshots (table)
Uses stage1_bridge_rows? No - reads from stage3_snapshots (downstream processed table) Uses transaction_events_raw? No - reads from processed table

​
3. Agent Focus Accounts (/api/v1/agent/focus-accounts)

Function: get_focus_accounts() Source Tables/Views:
  • payroll-bi-gauntlet.payroll_analytics.agent_churn_metrics_v (view) - Lost businesses
  • payroll-bi-gauntlet.payroll_analytics.agent_report_v5 (view) - Impact metrics (lost commission, lost employees)
Uses stage1_bridge_rows? No - reads from analytics views (agent_churn_metrics_v, agent_report_v5) Uses transaction_events_raw? No - reads from analytics views

​
4. Commissions (/api/v1/commissions)

Function: get_commissions() Source Tables/Views:
  • payroll-bi-gauntlet.payroll_processed.stage3_snapshots (table)
Uses stage1_bridge_rows? No - reads from stage3_snapshots (downstream processed table) Uses transaction_events_raw? No - reads from processed table

​
Summary

Dashboard endpoints do NOT directly read from:
  • transaction_events_raw (intake pipeline raw table)
  • stage1_bridge_rows (intake pipeline bridge table)
Dashboard endpoints read from:
  • stage3_snapshots (processed table - downstream from stage1)
  • ceo_snapshot (analytics view - aggregated from processed data)
  • agent_report_v5 (analytics view - aggregated from processed data)
  • agent_churn_metrics_v (analytics view - aggregated from processed data)
  • stage1_snapshots (raw table - but only for distinct counts in CEO metrics)
Drift Risk: None - dashboards read from downstream processed/analytics views, not from intake pipeline tables. The intake pipeline (transaction_events_raw → stage1_bridge_rows) feeds into downstream processing that populates these views.