Getting Started with Fleety
Fleety is an AI-powered customer support platform that helps you provide instant, intelligent support to your users. Add chat widgets and ticketing systems to your website with just a few lines of code, powered by advanced RAG (Retrieval-Augmented Generation) technology that learns from your documentation.
Why Fleety?
AI-Powered Responses
Advanced AI understands context and provides accurate answers based on your documentation.
5-Minute Integration
Copy-paste components into your app. No complex setup, no backend required on your end.
Secure & Anonymous
Short-lived tokens, origin validation, and optional custom authentication to prevent abuse.
Knowledge Base RAG
Upload your docs as ZIP files. AI automatically retrieves relevant context for each query.
Quick Start (3 Steps)
Create Your Project
After signing up, create a new project from your dashboard. You'll receive a unique project_id that identifies your integration.
Navigate to: Dashboard → Projects → Create Project
Save your project_id - you'll need it for integration!
Configure Project Settings
Set your allowed origins to prevent unauthorized use of your project:
https://yourdomain.com
https://www.yourdomain.com
http://localhost:3000 # For developmentOnly requests from these origins will be accepted by the Fleety API.
Integrate Chat or Tickets
Choose which widget you want to add to your site:
Important Concepts
🎟️ Anonymous Sessions
Chat widgets use short-lived anonymous tokens (5-minute TTL) that auto-renew. Users don't need accounts to get support. This is managed automatically by the components.
🛡️ Custom Authentication (Optional)
Want to restrict access to logged-in users only? You can wrap the widgets in your own auth guards and pass user context:
<!-- Only show to authenticated users -->
{#if $currentUser}
<SupportChat projectId="your-id" />
{/if}📊 RAG Knowledge Base
Upload documentation (ZIP with .txt, .md, .html, .csv, .json files) to your project. Fleety automatically chunks, embeds, and stores them in a vector database. When users ask questions, relevant chunks are retrieved and injected into the AI context.