Dashboard & Analytics

The Virsyn Voice AI dashboard system provides real-time and historical analytics across all call activity. Seven dedicated edge functions power KPI cards, trend charts, conversion funnels, campaign comparisons, call quality metrics, time-of-day heatmaps, and objection analysis.

Architecture

Raw Calls ──► process-enrichment-queue (AI analysis) ──► call_stats_daily (pre-aggregated)
                                                              │
               ┌──────────────────────────────────────────────┤
               │              │              │                │
         dashboard-      dashboard-    dashboard-       dashboard-
         overview       timeseries      funnel          campaigns
                                                              │
               ┌──────────────────────────────────────────────┘
               │              │              │
         dashboard-      dashboard-    dashboard-
          quality         heatmap      objections
                          (raw calls)  (raw calls)

• 5 endpoints query the pre-aggregated call_stats_daily table for fast lookups.
• 2 endpoints (heatmap, objections) query the raw calls table for hour-level granularity and JSONB array unnesting.

Endpoints

Function	Data Source	Description
dashboard-overview	call_stats_daily	Summary KPIs, disposition breakdown, quality metrics, tool usage
dashboard-timeseries	call_stats_daily	Daily aggregated metrics for trend charts
dashboard-funnel	call_stats_daily	5-stage conversion funnel with step-to-step rates
dashboard-campaigns	call_stats_daily + v_call_batch_summary	Active batch progress and campaign-level stats
dashboard-quality	call_stats_daily	Call quality analytics with daily trends
dashboard-heatmap	calls	Best-times-to-call day/hour distribution
dashboard-objections	calls	Top objections and buying signals analysis

Authentication

All endpoints require a Supabase JWT Bearer token. The backend resolves the user's agency_id automatically -- you never pass it as a parameter.

const response = await fetch(`${SUPABASE_URL}/functions/v1/<function-name>?params`, {
  headers: {
    'Authorization': `Bearer ${session.access_token}`,
    'Content-Type': 'application/json',
  },
});

Common Query Parameters

All endpoints share these filters:

Parameter	Type	Default	Description
period	string	7d	today, 7d, 30d, 90d, custom
start_date	string	--	ISO date YYYY-MM-DD (required when period=custom)
end_date	string	--	ISO date YYYY-MM-DD (required when period=custom)
campaign_id	UUID	--	Filter to a specific campaign
client_id	UUID	--	Filter to a specific client

note

The heatmap endpoint only supports 7d, 30d, 90d periods (no today or custom). The campaigns endpoint defaults to 30d instead of 7d.

Error Handling

All endpoints return errors in a consistent format:

{
  "error": "Human-readable error message",
  "details": "Technical detail (only in 500s)"
}

Status	Meaning
401	Missing or invalid Bearer token
403	User exists but has no agency_id
405	Wrong HTTP method (must be GET)
500	Server error (check details)

Data Freshness

Source	Freshness
call_stats_daily	Updated by trigger when call analysis completes + cron every 6h for last 7 days
v_call_batch_summary	Real-time view (queries on read)
Raw calls table	Real-time (used by heatmap and objections endpoints)

Next Steps

• Frontend API Guide -- TypeScript helpers, page layout mapping, and code examples