Giving your AI more context
How to ground your AI product in your knowledge, give it live web access, and connect it to your business tools.
Your AI Product is capable by default. These features make it specifically yours — grounded in your expertise, connected to live data, and wired into your business tools.
Three ways to extend your AI
Reference Material
Paste your pricing, methodology, FAQs, or any background context directly into the AI Chat settings. Claude has it available for every conversation — free, private, always on.
Web Search
Claude can search the web natively, without any additional nodes or configuration. It activates automatically when a conversation needs current data.
Integrations
Use webhooks to send completed response data to any CRM, email platform, or tool. For deeper two-way connections, add an MCP node to give Claude live access to your tools during conversations.
File uploads during conversations
If your product needs to analyse content your end user brings — a CV, a brief, a website URL — you can add a file upload or URL question type in your AI Chat node's Collection Goals. Your end user uploads their file mid-conversation, and Claude reviews it as part of the session.
This is different from Reference Material: Reference Material is your knowledge (always present), file uploads are your end user's content (session-specific).
The MCP node
The MCP node connects your product to any MCP-compatible server. Claude can call that server's tools during a conversation — looking up live CRM data, checking availability, writing back records — or after, based on instructions in the system prompt.
Read the full MCP node guide →
Most products work well with just the AI Chat node and an outcome. Add Reference Material when you want Claude to know your specific methodology and pricing. Use the webhook when you want leads routed automatically after completion.