Generating Docs and Updating Airtable
To turn the AI output into a Google Doc, set up the workflow so it runs in a loop, processing multiple ASINs one by one. Each iteration creates or updates a document and then writes the status and document link back to Airtable. To make this work, add the following new nodes:
- Google Docs: create a document;
- Google Docs: update the document;
- Split In Batches (or Loop Over Items) with batch size 1;
- Airtable: Update the record with the final Complete status;
- Set / Edit Fields to build a clickable Google Doc URL.
Creating and Updating the Google Doc
To generate a new Google Doc for each ASIN and fill it with the AI summary, add two Google Docs nodes in sequence.
-
Google Docs: create a document:
- Drive: My Drive;
- Folder: SaaS reports;
- Title: summary and tips (temporary title).
-
Google Docs: Update a document:
- Document ID/URL: drag from the Create a document output;
- Text: drag the output of the AI Agent (the LLM's summary).
Execute the nodes, open Google Drive, and confirm that the created document contains the AI-generated text.
If the AI output isn't mapped into the Text field, Google Docs throws an insert text must be… error. Make sure the LLM result is mapped correctly, then re-run the node.
Batching and Looping the Workflow
To prevent multiple ASINs from running at once, batch the flow so it processes items one by one.
-
Airtable search records:
- Limit: 10 (or the number you prefer);
- This node stays outside the loop.
-
Split In Batches (or Loop Over Items):
- Batch size: 1.
Connect the loop as follows:
Search Records → Split In Batches → (rest of the workflow) → back to Split In Batches
Keep the first Airtable Update Record inside the loop to mark each row as In Progress before processing. Run the workflow and watch items execute one at a time.
Why this shape works Marking all rows as In Progress at once can cause conflicts. A batch size of 1 keeps processing clean and sequential.
Giving Each Document a Unique Name
If all generated documents have the same title, include the ASIN in the file name. In Create a document → Title, append the ASIN dynamically:
summary and tips - {{$json.asin}}
Re-run the workflow to confirm that each file name now includes the correct ASIN.
Writing the Results Back to Airtable
After the Google Doc is updated, finish by writing the results and document link back to Airtable.
-
Set / Edit Fields (to build a clickable Google Doc URL):
- Field:
doc_url; - Value (expression):.
- Field:
https://docs.google.com/document/d/{{$node["Create a document"].json["id"]}}/edit
-
Airtable: update record:
-
Match on: ID (drag the record ID surfaced earlier);
-
Fields to update:
status→ Complete;- script → LLM output (same text you inserted into the doc);
doc_link→ thedoc_urlvalue created above.
-
Execute the nodes and verify in Airtable that the row now shows Complete, along with a clickable link that opens the corresponding Google Doc.
Takk for tilbakemeldingene dine!
Spør AI
Spør AI
Spør om hva du vil, eller prøv ett av de foreslåtte spørsmålene for å starte chatten vår
Awesome!
Completion rate improved to 4.17
Generating Docs and Updating Airtable
Sveip for å vise menyen
To turn the AI output into a Google Doc, set up the workflow so it runs in a loop, processing multiple ASINs one by one. Each iteration creates or updates a document and then writes the status and document link back to Airtable. To make this work, add the following new nodes:
- Google Docs: create a document;
- Google Docs: update the document;
- Split In Batches (or Loop Over Items) with batch size 1;
- Airtable: Update the record with the final Complete status;
- Set / Edit Fields to build a clickable Google Doc URL.
Creating and Updating the Google Doc
To generate a new Google Doc for each ASIN and fill it with the AI summary, add two Google Docs nodes in sequence.
-
Google Docs: create a document:
- Drive: My Drive;
- Folder: SaaS reports;
- Title: summary and tips (temporary title).
-
Google Docs: Update a document:
- Document ID/URL: drag from the Create a document output;
- Text: drag the output of the AI Agent (the LLM's summary).
Execute the nodes, open Google Drive, and confirm that the created document contains the AI-generated text.
If the AI output isn't mapped into the Text field, Google Docs throws an insert text must be… error. Make sure the LLM result is mapped correctly, then re-run the node.
Batching and Looping the Workflow
To prevent multiple ASINs from running at once, batch the flow so it processes items one by one.
-
Airtable search records:
- Limit: 10 (or the number you prefer);
- This node stays outside the loop.
-
Split In Batches (or Loop Over Items):
- Batch size: 1.
Connect the loop as follows:
Search Records → Split In Batches → (rest of the workflow) → back to Split In Batches
Keep the first Airtable Update Record inside the loop to mark each row as In Progress before processing. Run the workflow and watch items execute one at a time.
Why this shape works Marking all rows as In Progress at once can cause conflicts. A batch size of 1 keeps processing clean and sequential.
Giving Each Document a Unique Name
If all generated documents have the same title, include the ASIN in the file name. In Create a document → Title, append the ASIN dynamically:
summary and tips - {{$json.asin}}
Re-run the workflow to confirm that each file name now includes the correct ASIN.
Writing the Results Back to Airtable
After the Google Doc is updated, finish by writing the results and document link back to Airtable.
-
Set / Edit Fields (to build a clickable Google Doc URL):
- Field:
doc_url; - Value (expression):.
- Field:
https://docs.google.com/document/d/{{$node["Create a document"].json["id"]}}/edit
-
Airtable: update record:
-
Match on: ID (drag the record ID surfaced earlier);
-
Fields to update:
status→ Complete;- script → LLM output (same text you inserted into the doc);
doc_link→ thedoc_urlvalue created above.
-
Execute the nodes and verify in Airtable that the row now shows Complete, along with a clickable link that opens the corresponding Google Doc.
Takk for tilbakemeldingene dine!