The Knowify AI Connector connects Knowify directly to AI tools like Claude and ChatGPT, so you can ask AI questions about your business and get real answers based on your actual job data in Knowify.
Once connected, your AI system can answer questions about your Knowify data on demand. Ask "which jobs lost money last quarter?" or "what's ready to invoice?" and your AI tool pulls the answer from your live Knowify workspace.
This guide walks you through how to use the AI Connector to connect Knowify to Claude and ChatGPT.
A few things worth noting:
- The AI Connector uses an MCP server to connect Knowify to your AI tools. MCP (Model Context Protocol) is a secure connection point that any AI tool supporting the open MCP standard can connect to.
- It's read-only. AI tools can query and analyze your data via the MCP but can't create, edit, or delete anything in Knowify. We will add write capabilities in a future update.
- It is available for users with admin permissions in Knowify only. To authorize their connection, the Knowify user will need to have Admin permissions.
- Your AI tool subscription is separate from Knowify. You'll need an active account with Claude, ChatGPT, or another MCP-compatible LLM to use this.
How to connect Knowify to Claude
- Sign in to Claude at claude.ai.
- Open Settings from your profile menu, then select Connectors in the sidebar.
- Scroll to the bottom of the list and click Add custom connector.
- Enter the following:
- Name: Knowify
-
URL:
https://assistant.knowify.com/api/v2/mcp
- Click Add. Claude will connect to the Knowify MCP server and redirect you to authenticate via OAuth.
- Sign in with your Knowify credentials when prompted. This links your Claude account to your Knowify account for data access.
- Open a new chat. Click the + icon in the lower-left of the chat box and toggle Knowify on for that conversation.
- Ask your first question to confirm the connector is working. For example:
- "Show me my five most recent open invoices."
- "Which active jobs are tracking over budget this year?"
- "Summarize my WIP position as of today."
- "Are any subcontractor COIs expired or about to expire?"
Claude should respond with data pulled directly from your Knowify account. If Claude says it can't access Knowify, check Settings → Connectors and make sure the Knowify connector is toggled on for that chat.
How to connect Knowify to ChatGPT
ChatGPT's MCP support lives behind a setting called Developer Mode. It's a one-time toggle, but it's required.
Step 1: Turn on Developer Mode
- Log in to ChatGPT at chatgpt.com.
- Click your profile icon in the top-right corner, then select Settings.
- Go to Apps & Connectors (may also appear as Connectors depending on your account type).
- Scroll to the bottom and click Advanced settings.
- Toggle Developer Mode on. ChatGPT will show a warning that custom connectors run third-party code — read it, accept, and continue.
Step 2: Add Knowify as a custom connector
- Back in Apps & Connectors, click Add custom connector (you'll only see this option after Developer Mode is on).
- Fill in the connector details:
-
Name:
Knowify -
Description:
Query my Knowify data(this is what ChatGPT will show in chat) -
MCP Server URL:
https://assistant.knowify.com/api/v2/mcp
-
Name:
- Click Create.
- ChatGPT will redirect you to sign in to Knowify. Use your normal Knowify login credentials.
- Authorize the connection when prompted.
Step 3: Enable Knowify in a chat
Custom connectors in ChatGPT have to be turned on per-conversation. To use Knowify in a new chat:
- Start a new conversation.
- Click the + icon in the message box and select Developer mode (or the Connected apps menu, depending on your version).
- Toggle Knowify on for this chat.
You're connected. Ask your first question.
How to connect Knowify to Grok
- Go to https://grok.com/connectors.
- Click New connector, then Custom.
- Fill in the connector details:
Name:Knowify
Server URL:https://assistant.knowify.com/api/v2/mcp - Sign in with your Knowify credentials to authenticate the connection.
- You're connected. Start a new conversation and ask your first question.
What you can ask
Anything you manage in Knowify is available to query via your LLM. Here are few starter categories and prompts:
Job performance
- "What's my current margin on all active jobs?"
- "Which jobs are over budget or projecting a loss?"
- "Show me estimated vs. actual costs on the Riverside job, by phase."
WIP and billing
- "Pull my current WIP — which jobs are overbilled and which are underbilled?"
- "Which jobs have unbilled work ready to invoice?"
- "How much retainage am I holding across active jobs?"
Cashflow
- "Which customers have invoices more than 30 days past due?"
- "What's my total AR right now?"
- "Give me an AP aging summary."
Service and subcontractors
- "How did my service business perform last month?"
- "Are any subcontractor compliance documents expired?"
- "What service visits are scheduled this week?"
Business pulse
- "Summarize my business this week — active jobs, anything over budget, overdue invoices, open change orders."
- "What's my total backlog — work under contract but not yet complete?"
Phrase questions the way you'd ask a colleague. Claude will ask clarifying questions if it needs to narrow things down.
Tips for getting better answers
- Be specific when it matters. Including a job name, date range, or job type gets you a more precise answer.
- Ask for a format. "Give me this as a table" or "summarize in three bullet points" both work.
- Keep the conversation going. Claude can follow up on prior answers. If it pulls a list of jobs at risk, you can ask "which of those has the highest labor overrun?" without restating the context.
- Pair Knowify with your other tools. Because the connector is built on the open MCP standard, it works alongside any other MCP-compatible tool you've connected — QuickBooks, your CRM, your email — in the same conversation.
Questions or issues? Contact us at support@knowify.com. We’re here to help!
Comments
0 comments
Article is closed for comments.