Audit logs

Every operation performed in the Verta platform can create an audit log. This includes both API calls and model predictions. The audit log contains the relevant information to help you identify access patterns and potential suspicious activity.

Below is an example of a log entry:

{
    "user": {
        "id": "123",
        "email": "[e-mail]"
    },
    "timestamp": "2006-01-02T15:04:05.999999999Z",
    "method": "exampleMethod",
    "requestHeaders": {
        "header1": ["value1"]
    },
    "requestPath": "/example",
    "request": {
        "exampleType": {
            "id": 123
        }
    },
    "requestBase64": "ZXhhbXBsZQo=",
    "response": {
        "exampleOtherType": {
            "id": 123
        }
    },
    "responseBase64": "ZXhhbXBsZQo=",
    "responseHeaders": {
        "header1": ["value1"]
    },
    "returnCode": 500,
    "errorCode": 500,
    "errorMessage": "sampleErrorMessage",
    "requestId": "uuid",
    "versionId": "v1",
    "namespace": "production",
    "modelInfo": {
        "modelName": "my model",
        "modelVersion": "v0",
        "endpointPath": "/example",
        "endpointId": 123,
        "stageId": 456,
        "buildId": 12,
        "revisionId": 45
    }
}

The logs are saved in a configured object store as compressed new-line separated json files.

Reach out to us at [e-mail] for any questions on how to consume the logs or to disable audit logs in your account.