Request Json Format

AlgoMarker Request Format

For complete details, see the AlgoMarker Spec.

Required Fields

A valid request to AlgoMarker should include the following fields:

• type: Set to "request".
• request_id: A unique string identifier for the request.
• export: Specifies the desired output. For standard predictions, use "prediction": "pred_0" to request the model prediction and store it in the "prediction" field of the response. You can also specify request to recieve explainabiltiy output if applicable for example by specifiying "explainability_output_field_name_for_your_control": "json_attr Tree_iterative_covariance".
• load: Set to 1 to indicate that patient data is included in the request.
• flag_threshold (optional): If your model supports configurable thresholds, specify the threshold name here.
• requests: An array of patient requests, each containing:
  • patient_id: An integer patient identifier.
  • time: The calculation date in YYYYMMDD format. Any data after this date will be ignored from score calculation.
  • data: Patient data, including:
    • signals: An array of signal objects, each with:
      • code: The signal name.
      • data: An array of data points, each with:
        • timestamp: An array of timestamps (e.g., [20250806]). Can be empty for static values.
        • value: An array of values (e.g., ["14.5"] or [82]). You can pass string/float, both OK.
      • unit (optional): The unit for the signal (e.g., "fL").

Example Signal Entry

{
    "type": "request",
    "request_id": "999ef0f4-1099-4178-8c86-ecbfac6578e2",
    "export": {
        "prediction": "pred_0"
    },
    "flag_threshold": "USPSTF_50-80$PR_03.000",
    "load": 1,
    "requests": [
        {
            "time": "20250806",
            "patient_id": "1",
            "data": {
                "signals": [
                    {
                        "code": "Hemoglobin",
                        "data": [
                            {
                                "value": [
                                    "14.5"
                                ],
                                "timestamp": [
                                    20250806
                                ]
                            }
                        ],
                        "unit": "fL"
                    },
                    {
                        "code": "Hematocrit",
                        "data": [
                            {
                                "value": [
                                    "32"
                                ],
                                "timestamp": [
                                    20250806
                                ]
                            }
                        ],
                        "unit": "fL"
                    },
                    {
                        "code": "RBC",
                        "data": [
                            {
                                "value": [
                                    "40000.5"
                                ],
                                "timestamp": [
                                    20230806
                                ]
                            },
                            {
                                "value": [
                                    "4.5"
                                ],
                                "timestamp": [
                                    20250806
                                ]
                            }
                        ],
                        "unit": "fL"
                    },
                    {
                        "code": "MCV",
                        "data": [
                            {
                                "value": [
                                    82
                                ],
                                "timestamp": [
                                    20250806
                                ]
                            }
                        ],
                        "unit": "fL"
                    },

                    {
                        "code": "GENDER",
                        "data": [
                            {
                                "value": [
                                    "Male"
                                ],
                                "timestamp": []
                            }
                        ]
                    },
                    {
                        "code": "BDATE",
                        "data": [
                            {
                                "value": [
                                    19780101
                                ],
                                "timestamp": []
                            }
                        ]
                    }
                ]
            }
        }
    ]
}