Quick Start Guide
Get your AI Agent live on your website in just a few simple steps! This quickstart guide will have you up and running in minutes.
Sign up on kaily.ai
Visit kaily.ai and create your account with your email and password, or use Google for one-tap login.
Create an AI Agent
Create an AI Agent — either:
- Choose a prebuilt template (like Chat with Website, Chat with PDF, or Customer Support), or
- Click "Build from Scratch" to make your own.
Customize your AI agent
Once your agent is created, customize it by:
- Setting up custom greetings and responses
- Configuring the appearance and behavior
Go to the Deploy section
Navigate to your agent's dashboard and click on the Deploy section.
Open the Web Widget tab
In the Deploy section, open the Web Widget tab, then click Embed Code to get your script.
Copy the embed code
Copy the script snippet shown under Embed Code.
Paste it into your website
Paste the embed code into your website — inside either the <head> or <body> tag.
<!-- Paste the embed code anywhere in your HTML -->
<script type="application/javascript">
(function(w,d,s,o,f,js,fjs){w[o]=w[o]||function(){(w[o].q=w[o].q||[]).push(arguments);};(js=d.createElement(s)),(fjs=d.getElementsByTagName(s)[0]);js.id=o;js.src=f;js.async=1;js.referrerPolicy = "origin";fjs.parentNode.insertBefore(js,fjs);})(window,document,"script","copilot","https://script.copilot.live/v1/copilot.min.js?tkn=YOUR_TOKEN_HERE");
copilot("init",{});
</script>
That's it!
Visit your site, click the chat widget, and start talking to your AI agent.
Next Steps
- Add data sources: Connect your website, PDFs/docs, Quip, Notion, Google Drive, or Git to ground the agent in your real content and reduce hallucinations. See Data Sources.
- Connect Workflows (Boltic): Trigger automated actions from conversations (e.g., create tickets, send emails, update CRM) using Boltic-powered workflows with conditions and triggers. See Workflows.
- Connect MCP: Enable your agent to safely call external tools and fetch live data via the Model Context Protocol (MCP), extending capabilities beyond its training. Kaily acts as an MCP client, allowing your AI agent to execute tasks based on real-time data from a configured MCP server. See MCP Server.
- Monitor conversations: Use the Inbox to review chats, identify gaps, and iterate on prompts, knowledge, and workflows to improve outcomes. See Inbox.