# Interpreters

An Interpreter processes a user input and matches it to an intent. Intents act as categories of user utterances. For example "What will the temperature be like today?" could be matched to a "Temperature" intent, whereas "Hello" could be matched to a "Greeting" intent.

Interpreters also allow you to configure your bot with a variety of functionalities depending on your needs such as more granular intent matching, source analysis and response matching, Intent Classification, third-party NLU integration and webhook API integration. By default, OpenDialog comes pre-configured with a default 'OpenDialog' interpreter which is used throughout your scenario.

Interpreters are broken down into three categories in OpenDialog: 

* **Those that connect with Language Services** (Semantic Intent Classifiers, Knowledge Services, or Intent Classifiers) 
* **Third-party NLU interpreters** (OpenAI, Azure, Amazon LEX, Google Dialogflow, and custom configurations) 
* **Webhook Interpreters** (for API integration) 

You can choose the best Interpreter for your scenario from the 'Interpret' landing page. 

{% hint style="success" %}
To create an interpreter: 

* Log in to your OpenDialog account 
* Select a workspace and click 'Manage scenarios' 
* Review which scenario you would like to add an interpreter in
* Select the scenario, and navigate to the sidebar 
* Select 'Interpret' 
* From the Interpret landing page, select the Interpreter that best suits your needs 
  {% endhint %}

{% hint style="info" %}
Additional steps are required to complete the integration set-up between OpenDialog, and a third-party NLU service. You will also need to create intents in the third-party service.
{% endhint %}

Multiple Interpreters can exist within a single scenario, and each Interpreter can fulfil a different purpose. Every Interpreter you create will be shown as a card on the 'Interpret' landing page. 

<figure><img src="https://3203401256-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FdP6xFXXyeBbibro1Ywnq%2Fuploads%2FHRGa7iHHRwjufCAIDmpc%2F2023-05-02_12-11-35.png?alt=media&#x26;token=1050986d-46c3-4607-acd3-502739a4a9db" alt=""><figcaption><p>Interpreters overview page</p></figcaption></figure>

<figure><img src="https://3203401256-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FdP6xFXXyeBbibro1Ywnq%2Fuploads%2FweuR2lfxSSw9nHTM8X3F%2F2023-05-02_12-11-15.png?alt=media&#x26;token=672c5412-6ab1-494d-8edd-511be89810f4" alt=""><figcaption><p>Adding a new interpreter page</p></figcaption></figure>

Once an Interpreter is available and set to active using the toggle switch, it can be [referenced in the conversation structure](https://docs.opendialog.ai/core-concepts/the-opendialog-workspace/scenarios/turns-and-intents#interpreter-and-confidence-level) at the component level.&#x20;