Calling a detector

Once you have created a detector, you can use it by calling it through our Python library.

An detector’s ID is listed on the dashboard. It will look something like this:

68211f82-da64-43f9-85de-35e073c4e3fd

After installing the Haize Labs library, a detector can be called with the following code:

Python

from haizelabs.client import HaizeClient

client = HaizeClient(api_key="<your_api_key>")

response = client.detectors.call(
    detector_ids=["68211f82-da64-43f9-85de-35e073c4e3fd"],
    messages=[{"role": "user", "content": "hello from haize labs!"}]
)

We get the following response back. Since our input was Hello world, none of the detectors were triggered.

Response

[
    DetectorCallResponse(
        content_id='478557227c764dc78426478199309824',
        detected=False,
        detector_id='7c14403c-77c1-4c91-bc6d-a2a9a8f86aad',
        score=0.0
    )
]

We can now use this response to filter out unwanted content and monitor content categories.