🔧 Prerequisites
Access to Make (Integromat) or Zapier with ability to create scenarios.
A Google Sheets file with interaction data (columns:
A: Email
B: Date
C: Name
D: Notes)
A folk API key with permissions to create interactions.
An OpenAI API connection in Make (to clean/format notes).
📝 Step 1. Set up Google Sheets module
Add a Google Sheets > Search Rows (Filter Rows) module.
Connect your Google account.
Configure it:
Spreadsheet: select your “Add interactions” sheet.
Sheet:
Interactions
.First row:
A1:CZ1
(with headers).Filter: only take rows where Notes (column D) exists.
Sort order: Ascending.
This fetches all rows with notes to be converted into interactions.
🔎 Step 2. Search for the person in folk
Add an HTTP > Make a request module.
Configure it:
Method: GET
URL:
https://api.folk.app/v1/people?filter[emails][like]={{Email}}&filter[fullName][like]={{Name}}&combinator=or
(replace
{{Email}}
with column A and{{Name}}
with column C mapping).Header:
Authorization: Bearer YOUR_FOLK_API_KEY
Parse response: Yes
This looks up the person in folk to link the interaction.
🤖 Step 3. Clean notes with OpenAI
Add an OpenAI > Create Chat Completion module.
Connection: your OpenAI API key.
Model:
o4-mini
(or GPT-4 if available).Input prompt:
Convert the following plain text into clean, semantic Markdown with headings and lists. Remove links and return only the Markdown as a single line with literal \n characters. Text to convert: {{Notes}}
(map
Notes
= column D).
This ensures the interaction content in folk is clean and well-formatted.
🗂 Step 4. Create the interaction in folk
Add another HTTP > Make a request module.
Configure it:
Method: POST
URL:
https://api.folk.app/v1/interactions
Header:
Authorization: Bearer YOUR_FOLK_API_KEY Content-Type: application/json
Body (raw, JSON):
{ "entity": { "id": "{{PersonID}}" }, "dateTime": "{{formatDate(Date; "YYYY-MM-DD[T]HH:mm:ss.SSS[Z]")}}", "title": "Meeting with {{Name}}", "content": "{{replace(OpenAI Output; "\""; "'")}}", "type": "☕️" }
Map fields:
PersonID
=id
from the folk search (Step 2).Date
= column B.Name
= column C.OpenAI Output
= content returned in Step 3.
This creates a ☕️ meeting interaction attached to the right contact in folk. You can change that icon if you want.
🚦 Step 5. Error handling
If folk search (Step 2) returns no match, you may want to Ignore or Route those rows for manual review.
You can add a Make “Ignore” module on the error path of Step 2.
✅ Step 6. Test and schedule
Run the scenario once manually to test with a few rows.
Once confirmed, schedule it to run periodically (e.g., every hour/day).
📂 Alternative: Importing a JSON Blueprint into Make
Instead of recreating the scenario manually, you can import the provided JSON file:
Download the file: Template Make - add interactions on folk.json
Go to Make > Scenarios.
Click the arrow next to Create a new scenario and choose Import Blueprint.
Upload the JSON file.
Once imported, Make will recreate the entire scenario with modules already connected.
Replace placeholders with your own credentials:
Google Sheets connection → select your own Google account.
folk API → replace with your own API key in the HTTP modules.
OpenAI connection → map to your existing OpenAI account in Make.
Save and run the scenario once to test.