Automated Design Proofing System for Church Communications Using Monday.com, Make.com, and ChatGPT

In the world of church communications, coordinating events, promotions, and printed materials across a small team can quickly become an overwhelming, manual process. As Communications Director at a church, I’ve built a fully automated system that streamlines the design proofing workflow using Monday.com, Make.com (formerly Integromat), ChatGPT, Gmail, and internal team workflows. This article walks through the exact process, step-by-step, for how I automated the design proofing process at our church—from collecting initial event info to getting approvals and feedback with almost zero manual intervention.

If you’re managing a creative team, this DIY solution can eliminate the chaos of manually forwarding proofs, waiting for email replies, parsing vague feedback, and updating project statuses. Let’s dive into how it works.

Overview of the Communications Workflow
The Problem with Manual Proofing
Tools Used in This System
Creating the Automated Proofing Workflow
- Event Submission and Design Assignment
- Triggering the Automation via Webhook
- Designing the Make.com Workflow
- AI Proofing via ChatGPT
- Email-Based Client Approvals
Best Practices and Tips
Summary

1. Overview of the Communications Workflow

Our church communications department handles everything from design and signage to event promotion and website management. With a team of four—myself (the Director), a designer, a web & apparel manager, and a printer—we manage campaigns for internal events by collecting event info from different departments, assigning design tasks, and creating promotional content.

We use Monday.com as our primary hub to manage all project-related tasks.

Sample event lifecycle:

Event owner submits a form with event details.
Communication team reviews and assigns subtasks:
- Design
- Web registration
- Printing
- Close/Fulfillment
Designer uploads proofs to Monday.
Event owner must approve or request changes.

The biggest bottleneck was the proofing phase, which previously required repeated manual steps and back-and-forth emails.

2. The Problem with Manual Proofing

Here’s what the manual workflow looked like before automation:

Designer uploads proof to Monday.
I download the proof PDF.
I manually write and send an email to the event owner.
They send an email reply approving or suggesting changes.
I update Monday’s status manually and notify the designer.

Multiply that by dozens of projects a month, and this process became a huge time sink. It was inefficient and error-prone.

3. Tools Used in This System:

Monday.com: Project and task management
Make.com: No-code automation platform
ChatGPT (with API Access): AI-based PDF proofing
Gmail: SMTP/IMAP integration for sending/receiving emails
Vector Store/ChatGPT File Analysis: For PDF parsing and intelligent proofreading

4. Creating the Automated Design Proofing Workflow

Now let’s walk through how I built the fully automated solution.

STEP 1: Event Submission & Subtask Assignment

Event owner fills out a form (linked to Monday.com).
Monday.com auto-fills fields like:
- Date, time, location
- Preferred design style
- Additional assets (logos, images)
Designer is assigned a “Design” subtask.
Once the design is ready, they upload their PDFs in the “Proof Delivery” column.
Status is changed to “Proof”.

STEP 2: Triggering the Automation via Webhook

Using Monday’s built-in automations:

When a subtask’s status changes to “Proof,” a webhook is triggered to Make.com.

💡 Pro Tip: Use guest users in Monday to avoid paying for full team accounts! A separate tutorial (linked in the original transcript) covers this.

STEP 3: Creating the Make.com Scenario

This Make.com scenario is the heart of the automation.

Key Components:

✔️ Initial Webhook Received
Triggers upon the proof task status change.
🧠 Monday.com Module – Get Item
Retrieves all relevant task data (event owner, files, dates, etc.).
🚥 Router + Filter
Ensures the automation only runs for “Design” subtasks by checking a status column (like “Notify Owner”).
🔗 Iterator + Text Aggregator
Collects all uploaded proof files (can be multiple) and creates URL links for email inclusion.
📧 Gmail Email Module
Sends a fully formatted email to the event owner:
- Greets staff member.
- Lists links to proofs.
- Displays event date/time.
- Requests either “Approved” or feedback via email reply.
🔍 The email also includes an internal tracking number (Monday task ID), crucial for future steps.

Example email snippet:

Hello [Staff Member],

Please review your event proof linked below:
- [PDF Link]

Event Date: December 12th
Time: 6:00 PM

Reply with “Approved” or list requested changes.
Item ID: #5821

STEP 4: AI Proofing Using ChatGPT

In parallel, another router path in Make.com sends the same PDFs to ChatGPT for automatic quality review.

How It Works:

📄 File is uploaded to a Vector Store.
🤖 ChatGPT reads the file and compares it against the event metadata:
- Is the date/time formatted correctly?
- Are there spelling or grammar mistakes?
- Is branding consistent?
📝 ChatGPT generates a comment and posts it in Monday.com:
- Flags date mismatches
- Highlights incorrect time formats
- Suggests fixes (e.g., formatting “5P” to “5:00 PM”)
🚮 After proofing, Make.com deletes the file from the vector store to prevent clutter and API overages.

STEP 5: Processing Email Replies from Event Owners

Here’s where it gets even cooler.

Make.com has a separate “mailhook” scenario:

📥 A designated Gmail account forwards all proof replies to a Make.com webhook.
📊 ChatGPT analyzes the email:
- Is the proof approved?
- Is it rejected?
- What’s the specific feedback?
- What is the Monday item ID?
🔁 Based on content, Make.com:
- Updates Monday status to either “Approved” or “Changes Needed”.
- Posts the feedback as a comment on the task.
- Notifies the designer automatically.

✅ All the event owner needs to do is reply to an email—Make.com and ChatGPT handle the rest.

Full Loop Recap:

Designer marks proof as ready.
Automation sends email + files to event owner.
ChatGPT auto-proofs files.
Owner replies with “Approved” or revisions.
ChatGPT + Make.com process the email, update Monday, notify designer.
No one on our team has to manually handle anything.

5. Best Practices & Tips

❗Always include a unique internal tracking ID in your emails for mapping feedback.
🧠 Train ChatGPT with formatting rules (e.g., always use “5:00 PM” not “5P”).
♻️ Use routers and filters generously to avoid false triggers.
🧹 Clean your Vector Store after every file upload.
🧪 Test each automation module independently before chaining them.

This DIY solution has transformed how our team handles proofs, ensuring speed, clarity, and accountability. Stay tuned for follow-up tutorials diving deeper into each module.