# Quickstart - Chat Agent

The Ebbot Chat platform is designed to handle complex automated flows and multi-channel integrations. However, you can deploy a functional AI agent in just a few moments by connecting your knowledge base and publishing the chat widget to your website. This guide will walk you through how to do this.&#x20;

## Requirements for launching your Chat agent

Before you begin the implementation, ensure you have the following in place:

* **Access to Ebbot** – If you don't have an Ebbot account with EbbotGPT enabled yet, reach out to us and we will help you out. [Get in touch with us](https://www.ebbot.com/book-demo).&#x20;
* **Knowledge source** – A website URL, CSV or other files for your chat agent's knowledge.&#x20;
* **Deployment environment** – Editing access to your website to publish the chat widget and make it available for your end users.&#x20;

## 1. Add Knowledge

The knowledge you want to give to your chat agent is uploaded and handled under EbbotGPT Knowledge in the sidebar menu. Make sure to create a data set and simply add knowledge sources to this data set. For example, scrape a public website. Check out the video below to quickly understand how you scrape a website and user it as a knowledge source.

{% file src="<https://2117387010-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F3rWESGvwA3vHJ3zNiAG1%2Fuploads%2FHHOaWq2hRpd7PaR7R2Uz%2FDemo%20-%20Data%20source%20setup%20-%20Web%20Scrape.mp4?alt=media&token=132a423e-879e-4e17-9440-87353deae59f>" %}

## 2. Create a Configuration

Once you have scraped your website or uploaded files for your Chat agent to use as knowledge, it is time to create a Configuration. Navigate to EbbotGPT Configuration in the platform to get started.

Give your Configuration a name, select the default model, and connect the Data Set containing the knowledge you just uploaded.

Next, define the Chat agent's Persona. You can select an available persona template and edit the placeholders to suit your specific use case.

You also have the option to allow the Chat agent to hand over to a human agent, you may disable this if you only intend to implement AI functionality for now. Finally, set this Configuration as the Default to complete the setup.

## 3. Style the Chat widget

Ensure your chat widget aligns with your organization's branding by navigating to Chat widget in the sidebar menu. Under General Branding & Layout, you can customize colors and upload your logo and AI agent avatar. Remember to click Save in the upper right corner to apply your changes.

## 4. Chat with the Chat agent

Before going live, verify that your chat agent is functioning correctly. Click the chat bubble icon in the upper right corner of the platform to open the widget as an end user would see it. Test the agent by asking questions and reviewing the provided answers and sources. Once you are satisfied that the responses match your knowledge base and configurations, you are ready for deployment.

## 5. Implement the Chat widget to your website

Navigate to Chat widget > Settings to find the unique JavaScript snippet. Copy and paste this script into the of your website to make the widget accessible to your end users. Once implemented, your chat agent will be live and ready to assist your visitors.

Visit the below page for more information about implementing the chat widget.

{% content-ref url="../../developer-resources/ebbot-chat/implement-chat-widget" %}
[implement-chat-widget](https://docs.ebbot.ai/ebbot-docs/developer-resources/ebbot-chat/implement-chat-widget)
{% endcontent-ref %}

## Need any help?

You can reach our support team at any time by clicking Support in the sidebar menu. Start a conversation with our Support AI agent, if further assistance is required, the agent will automatically create a support ticket for our team to review and resolve.
