Our Implementation Engineers (IEs) will set up the model for you and provide you with a Bearer Token. This token is crucial as it allows you to access the API securely, ensuring safe communication between your systems and ours. It is crucial to keep this token safe and secure at all times, as it grants access to your Triage API.
Using the API
Send a POST request to
https://api.forethought.ai/api/predict
with the following structure:
{
"model_name": "Name of the model",
"text_a": "Subject or title",
"text_b": "Detailed description of the request"
}
Authentication
Authentication is required and done through a Bearer Token sent through the request header. Forethought will provide the token for the model via an expirable paste.
Use Case
{
"model_name": "Test",
"text_a": "Connectivity coverage",
"text_b": "I have an issue with connectivity coverage"
}
- "model_name" is the name of the model
- "text_a" is a text field that represents the subject or title
- "text_b" is a test field that represents the body or description of the request (or support case)
API Response
The Triage API will analyze your request and return a structured response that includes:
- "predicted_value" is the main category or issue identified
- "made_prediction" indicates if the problem meets the confidence threshold
- "prediction_confidence" is a score between 0 and 1, reflecting how confident the model is about its prediction
- "top_predictions" is a list of potential issues/topics ranked by confidence
Example Response
{
"predicted_value": "connectivity_coverage",
"made_prediction": true,
"prediction_confidence": 0.605250835,
"top_predictions": [
[
0.605250835,
"connectivity_coverage"
],
[
0.138640553,
"hardware_software"
],
[
0.121609338,
"calling"
],
[
0.107648298,
"texting"
],
[
0.026851017,
"number"
]
]
}