AI Spend Analytics

COST INTELLIGENCE

Track LLM token costs across agents, models, and departments. ROI calculator shows governance savings vs. manual equivalent. Budget alerts and trend analysis.

Real-timeCost Tracking
ROISavings Calculator
DeptBudget Breakdown

See it in action

custodex.vercel.app
COST INTELLIGENCE — Custodex Dashboard

Every token. Every model. Every dollar.

Automatic cost event creation from telemetry when token counts and model IDs are present. Track input tokens, output tokens, and estimated cost per action. Aggregate by time window (24h, 7d, 30d), agent, model, department, or action type.

Prove the value of AI governance.

Configure human hourly rate and minutes-per-action to calculate manual equivalent cost. Compare against actual AI spend to show savings percentage. Demonstrate governance ROI: prevented incidents, blocked unauthorized actions, and compliance automation savings.

// ROI calculation
{
  "windowMs": 604800000,
  "aiCost": "$1,247.50",
  "manualEquivalent": "$18,750.00",
  "savings": "$17,502.50",
  "roiPercent": 93.3,
  "eventCount": 15000,
  "avgCostPerAction": "$0.083"
}

Group costs by team. Set limits. Get alerts.

Breakdown view groups costs by department, agent, model, or action. Identify top spenders and cost outliers. Set budget thresholds with alerts when spend approaches limits. Trend analysis shows cost trajectory over time.

What's included in
cost intelligence

01

Automatic cost tracking from telemetry token counts

02

Per-model pricing with input/output token differentiation

03

ROI calculator with manual-equivalent comparison

04

Breakdown by department, agent, model, or action

05

Budget threshold alerts for cost control

06

Trend analysis with 24h, 7d, and 30d windows