Monitor AI credit consumption, usage trends, and quality metrics with built-in analytics.

AI Usage Monitoring

LintPDF provides detailed usage analytics for AI inspections, including per-feature breakdowns, trend analysis, and spending tracking.

Viewing Usage

Fetch AI usage data for your tenant:

curl -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.lintpdf.com/api/v1/ai/usage?start_date=2026-03-01&end_date=2026-03-16"

Filtering Options

Parameter	Description	Example
`start_date`	Start of date range (ISO 8601)	`2026-03-01`
`end_date`	End of date range (ISO 8601)	`2026-03-16`
`category`	Filter by AI category	`barcode_detection`
`feature`	Filter by specific feature	`barcode_decode`

Response

{
  "usage": [
    {
      "date": "2026-03-15",
      "category": "barcode_detection",
      "feature": "barcode_decode",
      "job_count": 42,
      "credits_consumed": 42,
      "total_cost": "5.04",
      "avg_processing_time_ms": 1250
    }
  ],
  "total_credits": 156,
  "total_cost": "18.72",
  "period": {
    "start": "2026-03-01",
    "end": "2026-03-16"
  }
}

Usage Trends (SPC)

The trends endpoint provides statistical process control (SPC) data for monitoring submission quality over time:

curl -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.lintpdf.com/api/v1/ai/usage/trends"

Response

{
  "trends": [
    {
      "date": "2026-03-15",
      "total_jobs": 50,
      "ai_jobs": 35,
      "total_findings": 120,
      "ai_findings": 45,
      "error_count": 2,
      "warning_count": 18,
      "info_count": 25,
      "avg_findings_per_job": 2.4,
      "pass_rate": 0.82
    }
  ],
  "spc": {
    "mean_findings": 2.1,
    "ucl": 4.8,
    "lcl": 0.0,
    "std_dev": 0.9,
    "out_of_control_dates": []
  }
}

The SPC (Statistical Process Control) data includes:

mean_findings: Average findings per job over the period
ucl / lcl: Upper and lower control limits (mean +/- 3 standard deviations)
std_dev: Standard deviation of findings per job
out_of_control_dates: Dates where findings exceeded control limits (signals quality shifts)

Admin Usage View

Admins can view AI usage across all tenants:

curl -H "X-Admin-Key: YOUR_ADMIN_KEY" \
  "https://api.lintpdf.com/api/v1/admin/ai/usage?start_date=2026-03-01"

This returns aggregated usage grouped by tenant, useful for billing reconciliation and capacity planning.

Spending Alerts

Configure Webhooks for proactive credit balance notifications:

curl -X POST https://api.lintpdf.com/api/v1/webhooks \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://your-app.com/webhook",
    "events": ["ai.credits.low", "ai.credits.depleted"]
  }'

Available AI-specific events:

Event	Description
`ai.credits.low`	Credit balance dropped below 20% of last package purchase
`ai.credits.depleted`	Credit balance reached zero — AI inspections will be skipped
`ai.circuit_breaker.open`	Vision circuit breaker tripped — Vision inspections temporarily unavailable
`ai.circuit_breaker.close`	Vision circuit breaker reset — Vision inspections available again

If a monthly_spending_limit is configured on your AI config, AI inspections are skipped when the limit is reached. Core engine checks continue normally. Monitor your spending via the credits endpoint:

curl -H "Authorization: Bearer YOUR_API_KEY" \
  "https://api.lintpdf.com/api/v1/ai/credits"

This returns your current balance, active packages, and month-to-date spending.

Best Practices

Set spending limits — Configure monthly_spending_limit to prevent unexpected charges
Monitor trends — Use the SPC endpoint to detect quality regressions early
Review by category — Filter usage by category to understand which AI features provide the most value
Archive usage data — Periodically export usage data for long-term analysis and compliance