Gmail RFI Tracker with OpenAI & Data Tables | n8n Workflow

Turn RFI emails into tracked, routed, and approved answers—automatically

This Gmail RFI Tracker with OpenAI & Data Tables n8n workflow monitors your RFI Gmail inbox, uses OpenAI to classify and extract RFI details (including text from attached PDFs), logs everything in n8n Data Tables, routes new questions to your design team, and sends timely replies back to the requester with PM approval when impacts are detected.

What this workflow does

• Listens for new RFI emails matching your configured Gmail search query in the RFI mailbox.
• Extracts PDF text from attachments (when present) and uses OpenAI to classify the message as:
  • a new RFI question,
  • a design-team answer, or
  • something that requires manual review.
• Matches RFIs to the correct project using project data stored in n8n Data Tables.
• Assigns RFI number and due date for new questions, then checks for duplicates against previously answered RFIs.
  • If duplicate: logs to the RFI log Data Table and replies with the prior answer via Gmail.
  • If new: logs to the RFI log Data Table, emails the RFI to the design team, and acknowledges receipt to the requester with the requested response date.
• Routes design-team responses by looking up the referenced RFI number in the RFI log; flags for review when it can’t be matched.
• Requests PM approval for impacts: when an answer indicates cost or schedule impact, it sends a Gmail approval request to the project manager before replying to the requester.
• Sends daily overdue reminders and digests to keep responses moving.

Use cases

• Automate RFI intake for construction/engineering teams receiving questions via a dedicated Gmail inbox.
• Reduce manual tracking by automatically numbering RFIs, detecting duplicates, and logging outcomes in Data Tables.
• Ensure cost/schedule responses are not sent without PM approval.

Technical details

• n8n nodes/integrations: Gmail, Data Table, OpenAI, if, switch, set, code.
• Data management: projects and RFI history stored in n8n Data Tables for lookup, deduping, and logging.
• Core logic: classification/extraction with OpenAI, conditional routing with if/switch, and structured updates via Data Tables.