What Analytics Does Cencori Provide?

Cencori's analytics dashboard gives you complete visibility into your AI infrastructure:

Request Volume: Track total requests and trends over time.
Cost Analysis: Detailed breakdown of spend by model and provider.
Performance: Monitor latency and success rates.
Security: Track and visualize security incidents (PII, injections).
User Insights: Analyze usage patterns by user ID or API key.

Key Metrics

Metric	Description	Goal
Total Requests	Number of AI requests processed.	Monitor growth
Success Rate	Percentage of successful vs error responses.	> 99%
Total Cost	Cumulative spend (Provider + Markup).	Stay within budget
Avg Latency	Time from request to response.	< 2s for UI tasks
Security Incidents	Count of blocked or flagged requests.	Minimize threats

Charts and Visualizations

Requests Over Time

Identify peak usage periods and long-term growth trends with our interactive line charts.

Cost Breakdown

Pie and bar charts show you exactly where your budget is going. Compare costs across OpenAI, Anthropic, and Gemini models.

Model Distribution

See which models are most popular in your application to help with model consolidation or migration decisions.

Security Threat Map

Monitor the distribution of security incidents over time to identify coordinated prompt injection attacks or PII leakage trends.

Filtering and Segmentation

Slice your data to find exactly what you're looking for:

Time Range: Filter by last 24h, 7 days, 30 days, or custom ranges.
Model/Provider: Compare performance and cost across different vendors.
API Key: Track usage per developer, environment (staging/prod), or application.
User ID: (Optional) If you pass user IDs, you can analyze behavior per user.

Common Use Cases

Performance Debugging

Find slow requests by filtering for high latency. Identify if specific models or provider regions are causing bottlenecks.

Capacity Planning

Use request trends to forecast future usage and ensure your organization has enough credits.

Security Auditing

Review spikes in security incidents to determine if they correlate with new deployments or external traffic shifts.

Accessing Analytics via API

Fetch metrics programmatically for your own internal dashboards using the Metrics API (/api/v1/metrics):

Codetext

// fetch-analytics.ts
const response = await fetch(
  'https://cencori.com/api/v1/metrics?period=24h',
  {
    headers: {
      'Authorization': `Bearer ${CENCORI_API_KEY}`
    }
  }
);
 
const { data } = await response.json();
// data includes request counts, cost, latency, and breakdowns by provider/model

Best Practices

Review weekly: Set a recurring task to check your cost and success rate trends.
Set Alerts: Configure low-balance and cost-threshold alerts to avoid surprises.
Track versions: Use different API keys for different versions of your app to compare their efficiency.
Share reports: Export analytics views to share with your security or finance teams.

Analytics Overview