Decision Simulation API

Simulate AI decision outcomes under various scenarios with comprehensive risk analysis and performance metrics.

Endpoint

POST /api/decision/simulate
                

Runs decision simulations for model validation and risk analysis

Authentication

Bearer JWT in Authorization header
Get authentication token

Request Parameters

Required Parameters

decision_model_id
String - UUID of pre-trained decision model
scenarios
Array of scenario JSON objects

Optional Parameters

analysis_type
String - 'sensitivity' or 'stress' (default: 'sensitivity')
iterations
Integer - Number of simulation runs (1-1000)

Example Request

POST /api/decision/simulate HTTP/1.1
Authorization: Bearer YOUR_JWT_TOKEN
Content-Type: application/json

{
  "decision_model_id": "d3c9f8a2-1e8e-4b6d-982a-64c788d010a4",
  "scenarios": [
    {
      "input_data": {
        "user_age": 35,
        "credit_score": 720,
        "income": 95000
      },
      "weights": {
        "risk_threshold": 0.7
      }
    }
  ],
  "analysis_type": "stress",
  "iterations": 500
}
                

Authorization:

Replace YOUR_JWT_TOKEN with your active authentication token obtained from /auth/token

Response Format

{
  "status": "success",
  "request_id": "req_5j21nk7q9p0lgs4392l4",
  "results": {
    "summary": {
      "mean_accuracy": 0.927,
      "risk_profile": "moderate",
      "convergence_rate": 0.78
    },
    "analysis": {
      "sensitivity": {
        "most_influential": ["credit_score", "income"],
        "impact_factors": {
          "credit_score": 0.42,
          "income": 0.31,
          "user_age": 0.12
        }
      },
      "stability": {
        "confidence_interval": {
          "lower": 0.89,
          "upper": 0.94
        },
        "variance": 0.021
      }
    }
  },
  "logs": [
    {
      "level": "info",
      "timestamp": "2025-08-25T18:30:45Z",
      "message": "Simulation completed with 500 iterations"
    }
  ]
}
                

Usage Notes

  • Each scenario must be less than 10KB in size
  • Rate limits: 100 requests/minute
  • Results are cached for 24 hours based on model ID

Need help with your simulations?

Our support team can help you design effective simulation scenarios and interpret the results.

Contact Technical Support