Skip to content
Haize Labs Docs

Creating a detector

In order to use a detector, it first has to be created and configured with content categories to detect.

Detectors can be created in the Haize Labs dashboard using the following steps:

  1. Login to the Haize Labs Dashboard and select the Detectors tab on the left sidebar.

  2. On the Detectors page, press the ‘Create Detector’ button.

  3. Configure your detector to detect the content you want.

    • Give a name to the detector (this is just for you to see).

    • Choose a detector type.

      • Natural language detectors can detect any content you can specify using natural language.

        • Ex: Information about nuclear centrifuges
        • Ex: Financial advice without a disclaimer
        • Ex: More than 5 emojis in a single sentence

      • Categorical detectors are preimplemented detectors made by Haize Labs to detect common categories of content.

      • Text matching detectors use regex patterns to find text content.

    • Configure your selected detector type.

  4. Press the ‘Create Detector’ button.

Your detector is now created! Detectors can be called using our Haize Labs Python library.

On the detectors home page, you’ll see that your detector has an unique ID similar to:

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

This ID can now be used to call your detector! Read calling a detector to learn how to use your detector to catch unwanted content.