Automating Proofing Workflows Between Monday.com, Make.com, and ChatGPT — A Deep Dive

ByMichael Fox

In the world of church communications and design requests, building automated systems that eliminate manual steps can dramatically increase efficiency and reduce errors. In this in-depth tutorial, we demystify a powerfully integrated system that uses Monday.com, Make.com, and ChatGPT to fully automate the design proofing and approval workflow within a church communication team — from automatically emailing proofs, processing AI and human feedback, and updating statuses inside Monday.com, to even closing out projects with file delivery — all without touching a thing.

Let’s break this system down from its triggers to its final status change, outlining each step in comprehensive detail so you can replicate or adapt it for your needs. Even if you’re not a seasoned developer, having an intermediate understanding of tools like Make.com (formerly Integromat), Monday.com APIs, and ChatGPT’s API functionality will help you fully grasp and deploy this automated proofing pipeline.

Overview

This workflow accomplishes three main goals:

  1. Send proofing email alerts and AI proof reads when a certain status is triggered in Monday.com.
  2. Process feedback from human stakeholders via email responses using ChatGPT, and automatically update project status accordingly.
  3. Deliver final assets via email and automatically close the task once the design is marked as “released”.

Part 1: Sending the Proofs Automatically When Status Changes

Step 1: The Trigger — Changing Status in Monday.com

Your project begins with a change in a “status” column in Monday.com (e.g., changing a design task to “Proof”). This triggers a webhook in Make.com.

✅ Prerequisites:

  • Set up a webhook within Make.com.
  • Familiarize yourself with Make.com Webhooks (see Make Documentation).

Step 2: Receiving Webhook in Make.com

The webhook receives the item ID from the Monday.com task status change. This ID is the key to fetching additional data.

Configure Your First Module: Get Item Details from Monday.com

  1. Add a new Monday.com module: “Get Item”.
  2. Authenticate using Monday.com’s API key.
  3. Use the ID from the webhook to fetch the entire item (which includes all column values).

🔍 Tip: Run the scenario once and inspect the bubble above the module to find the exact column you need, such as the proof owner’s email, file array, or a specific status label.

Step 3: Filter the Design Subtask

Not all items need processing. A filter is set to verify a specific design subtask by evaluating the content of a known column (Column 13). Only if this column contains the value “design” will the scenario proceed.

🛠 How-to:

  • Filter: columnValues[13].text EQUALS "design"

Part 2: Route One — Email Ownership Proof

Step 4: Iterate Through Files

To email all attached files as hyperlinks:

  1. Add an Iterator module pointing at the file array fetched from Monday.com.
  2. Add a Text Aggregator to build HTML anchor (<a href>) tags for each file using their name and URL.
  3. Aggregate all to one HTML-formatted block.

Step 5: Send Email via Gmail

Configure Gmail module:

  1. From: A pre-defined sender (use this same account for reply parsing later).
  2. To: Pull email from Column 10 (proof owner).
  3. Subject: Proof - {{Item Name}}
  4. Message Body:
    • Greet with the person’s name (Column 9).
    • List the design files as hyperlinks using aggregated HTML.
    • Include project date and time from Monday.com.
    • Append internal tracking with the Monday.com item ID (e.g., “monday_id: 123456”).

Part 3: Route Two — AI Proof with ChatGPT

Step 6: Iterate Through Assets Again

Same as before, this iterator pulls the assets attached to the item.

Step 7: Download Files

For each file:

  1. Download using the file ID via Monday.com modules.

Step 8: Upload to ChatGPT Vector Store

💡 Prerequisite:

  • Set up an Assistant and Vector Store in OpenAI via the developer portal.
  • Add an API Key to Make.com’s HTTP module.

Steps:

  1. Use an HTTP module to call POST /v2/files and POST /v2/vector_stores/{store_id}/files.
  2. Upload the PDF with proper JSON headers.
  3. Store the file in the design proof Vector Store.

Step 9: Message ChatGPT Assistant

Send a completion message to your assistant:

  • Instruct the bot to analyze and proof the document.
  • Include expectations like checking date/time format, spelling, grammar, etc.

The result from this analysis is captured into a downstream variable.

Step 10: Create Update in Monday.com

Post an update on the item using the captured AI message:
“This is the automated AI response: {{ChatGPT Result}}”

Step 11: Delete File from Vector Store

To prevent proofing old files:

  1. Use HTTP DELETE with the file ID and vector store ID in OpenAI’s API.
  2. Add relevant headers and authorization key.

Part 4: Responding to Proof Feedback Via Email

Step 12: Setup Gmail to Forward Replies

  1. Use a specific Gmail address to send original proofs.
  2. Set Gmail rule: All replies forwarded to a mailhook address (created in Make.com).

Tip: Configure Gmail’s filters appropriately to prevent unintended triggers.

Step 13: Capture Email in Make.com

Your mailhook triggers a new scenario upon receiving the email.

Step 14: Extract Monday.com ID from Email

Use ChatGPT’s text completion to:

  • Look at the last line for text like: monday.com proof: 123456.
  • Return only the ID or the word SKIP.

Filter out invalid or spam emails using this logic.

Step 15: Store the ID for Use

Put the returned ID into a text aggregator or variable for continuity.

Step 16: Analyze Proof Feedback Email

Send the email to ChatGPT again with formatting instructions:

  • Evaluate if the response includes approval (e.g., “I approve”, “Looks good”).
  • If changes are needed, extract those and output.

Step 17: Route Logic Based on Feedback

APPROVED:

  • Check for internal proof status to see if other team members approved.
  • If “No Changes Needed”, set status to Proof Approved.
  • If “Changes Needed”, set status to Proof Revisions.

NOT APPROVED / Changes Needed:

  • Automatically set status to Proof Revisions.
  • Add update with quoted message from the person who made the changes.

Part 5: Final Delivery and Task Closeout

Step 18: Designer Marks Task as “Released”

A Monday.com automation:

  • Triggers webhook when a design reaches “released”.
  • Auto-creates sub-task: “Close It”

Step 19: Handle Close It Task in Make.com

Webhook passes to a Make.com scenario. Use filter:

  • Column 13 must say “Close”

Step 20: Email Final Assets

  1. Iterate files again.
  2. Aggregate links using <a href>.
  3. Send final wrap-up email to the owner with downloadable links.

Step 21: Close Task in Monday.com

  • Set subtask status to Done

Full Automation — No Manual Interaction Required

With this structure:

  • Proofs go out automatically.
  • Human and AI feedback is processed.
  • Files are delivered and archived.

All without ever manually opening Monday.com, writing an email, or tracking statuses.

Similar Posts

Automate Your Lyrics: Syncing ProPresenter with Playback using MIDI on a Mac!

Automate Your Lyrics: Syncing ProPresenter with Playback using MIDI on a Mac!

This guide details how to integrate ProPresenter with Multitracks.com Playback on a Mac Mini to automate syncing lyrics with audio tracks. It covers equipment requirements, and provides a step-by-step setup process from preparing MIDI Network Session to final testing for live event production enhancement.