Hi @anawilliam850 

Jennifer here from Intercom

Fin is an AI bot powered by a mix of models, including OpenAI's GPT-4, alongside Intercom’s own proprietary AI technologies.

You can learn more about Fin here- https://www.intercom.com/help/en/articles/7120684-fin-explained 😊

Hope it helps

Integrating Intercom with a chatbot powered by GPT-3.5, like ChatGPT, can enhance the customer support and engagement capabilities of your Intercom-powered chat system. To achieve this integration, you'll typically follow these steps:

1. Access GPT-3.5 API: To interact with Chat GPT, you'll need access to the GPT-3.5 API. Ensure you have an API key or access to the API provided by the platform or service that hosts the GPT-3.5 model.

2. Set Up an Intercom App: If you don't already have an Intercom app, create one in your Intercom developer account. This app will serve as the bridge between Intercom and the GPT-3.5 model.

3. Configure Intercom Webhooks: Use Intercom's webhooks to send user messages from Intercom to your GPT-3.5-powered chatbot. You'll need to configure webhooks to trigger specific actions when certain events occur in Intercom, such as when a new conversation is initiated or when a message is received.

4. Develop the Chatbot Logic: Build a logic layer for your chatbot that communicates with the GPT-3.5 API using the API key. When Intercom sends a user's message to your webhook, your logic layer should forward that message to the GPT-3.5 API and receive a response.

5. Process GPT-3.5 Responses: After receiving a response from the GPT-3.5 model, you can process it as needed. You may want to format it, extract relevant information, or perform additional actions based on the response.

6. Send Responses Back to Intercom: Once you have processed the GPT-3.5 response, send it back to Intercom using their API. This will allow Intercom to display the response in the chat conversation with the user.

7. Testing and Quality Assurance: Thoroughly test the integration to ensure that messages flow correctly between Intercom and the GPT-3.5 chatbot. Verify that the responses are accurate and meet your expectations.

8. Deploy the Integration: Once you are satisfied with the integration's performance, deploy it to your production environment.

9. Monitoring and Maintenance: Continuously monitor the integration to ensure it operates smoothly. Be prepared to make updates and improvements as needed, especially if you encounter issues or receive feedback from users.