Vectorize Agents
Vectorize Agents are prebuilt, managed interfaces that connect your data pipelines to users through purpose-built interfaces. Deploy Chat Apps, embed Chat Widgets, or connect development tools through the Model Context Protocol (MCP).
Agent Types
Chat App
A standalone chat application with its own URL. Provides a full-featured conversational interface with user management, authentication, and admin controls for internal teams or external customers.
Chat Widget
An embeddable chat component for websites and applications. Add conversational AI to any web page with a single line of code, complete with customization and domain-based security.
MCP Agent
Model Context Protocol integration for AI desktop tools. Connect Claude Desktop, Cursor, and other MCP-compatible applications directly to your Vectorize pipelines without building a UI.
When to Use Which Agent
| If you need... | Choose... | Key features |
|---|---|---|
| Customer-facing chat on website | Chat Widget | One-line embed, domain security, customization |
| Internal knowledge base | Chat App | User management, SSO, multi-model support |
| IDE code assistance | MCP Agent | Native tool integration, code-aware context |
| Public documentation bot | Chat Widget | No auth required, easy embedding |
| Employee support portal | Chat App | Advanced auth, usage tracking, admin panel |
Next Steps
- Create a Chat App - Deploy internal AI assistant
- Embed a Chat Widget - Add AI chat to website
- Set up MCP Agent - Connect development tools
- Create Your First Pipeline - Set up data processing