If you want to summarize PDF with ChatGPT and save the result in Notion, one no-code Make scenario can do the whole job for you. Drop any PDF into a watched inbox or folder, ChatGPT writes a short structured summary, and the summary lands in Notion as a searchable page — no re-reading, no code.
What you need
- A Gmail inbox or Google Drive folder that receives PDFs
- A free Make account
- An OpenAI account
- A Notion database with properties for title, summary, and key points
Disclosure: Some links below are affiliate links, meaning StackScale may earn a commission at no extra cost to you.
Get started with Make here: https://www.make.com/en/register?pc=stackscale&affiliateSource=devto
The workflow in four modules
- Gmail watch emails or Google Drive watch files — the trigger, fired when a new PDF arrives.
- Extract text from the PDF — pulls the document text out so ChatGPT can read it.
- OpenAI chat completion — turns the text into a short, fixed-structure summary.
- Notion create a database item — saves one page per PDF, filed in arrival order.
Step 1: Set up the trigger
In Make, create a new scenario and add a trigger that fires when a new PDF arrives — a watched Gmail inbox with PDF attachments, or a Google Drive folder you drop files into. Pick the one your documents already land in, so nothing about your workflow changes.
Step 2: Extract the text from the PDF
Add a module that extracts the text from the PDF, or a Drive download step that hands the file over as data. ChatGPT reads text, not files, so this step is what makes the whole flow possible.
Step 3: Ask ChatGPT for a fixed-structure summary
Add the OpenAI module, create a chat completion, and give it a strict instruction so every summary comes back in exactly the same shape. Use a prompt like this:
Read the document below and return a summary with exactly three parts:
1. One-paragraph overview
2. Five key points
3. Action items with dates
Keep the summary under 200 words.
A fixed structure means every summary looks the same in Notion, and you can scan any of them in seconds.
Step 4: Save the summary in Notion
Add the Notion create-a-database-item module and map the fields like this:
{
"title": "{{document name}}",
"summary": "{{openai summary overview}}",
"key_points": "{{openai summary key points}}",
"action_items": "{{openai summary action items}}"
}
One page per PDF, filed in the order they arrive. Set the database up once with those properties, and the scenario fills it forever.
Step 5: Test with dummy data first
Upload a made-up report — a fake project update, a fake set of numbers — and run the scenario once. Check the page appears in Notion with the right summary. If ChatGPT misses a key point, tighten the prompt and run it again.
The morning summary habit
Every morning, open the Notion database first and read the newest summaries before you open a single PDF. The automation does the reading; you do the deciding.
Watch the full tutorial on YouTube: https://youtu.be/8KHPThr2FuY — and subscribe to StackScale for a new automation tutorial every day.
Top comments (0)