AI-Powered SaaS Usage Review and Alerting Agent

This Apify actor analyzes SaaS usage logs or API-provided JSON data to detect anomalies such as sudden drops or spikes in usage. It provides automated alerts with AI-generated commentary explaining the potential implications of detected anomalies.

Features

• Accepts either user-uploaded usage logs or fetches usage data from a provided API endpoint.
• Performs simple anomaly detection based on usage value deviations.
• Generates human-friendly AI interpretation comments for each anomaly.
• Pushes structured results including anomaly details and explanations.

Input

Provide input as JSON with one of the following fields:

• usage_logs: An array of objects each containing at least timestamp (string) and usage (numeric) or value (numeric).
• api_endpoint: URL string to fetch usage data JSON array.
• (Optional) api_headers: object with HTTP headers for the API request.

Example:

{
  "usage_logs": [
    {"timestamp": "2024-05-01T00:00:00Z", "usage": 120},
    {"timestamp": "2024-05-02T00:00:00Z", "usage": 30},
    {"timestamp": "2024-05-03T00:00:00Z", "usage": 115}
  ]
}

Output

Pushed to dataset is a JSON object with:

• status: "success" or "error"
• total_points: number of data points analyzed
• anomalies_count: number of anomalies detected
• anomalies: list of anomaly details, each including:
  • data_point: original data point of anomaly
  • commentary: AI-generated explanation

Usage

Deploy to Apify and run with your input. You can schedule runs or trigger via webhooks as needed.

Notes

• This Actor uses basic statistical anomaly detection suitable for straightforward usage metrics.
• For complex needs consider integrating with more advanced AI or ML anomaly detection models.