Hello everyone, I’m new to Make 
, so I don’t know much about it, but I’ll try.
I am creating a Make workflow where I need to:
Extract a number of files (images and PDFs) from a Tally form.
Perform OCR on each file using Google Cloud Vision.
Put all the extracted text in a single result.
Push that single text once to a Gemini module for further analysis.
So far, the iterator correctly processes each file, and the OCR and aggregation steps function as expected. However, the Gemini module still runs once for each file instead of just a single time after aggregation.
**
**
Here’s what I’ve tried:
Adding a filter on bundle.orderPosition = 1 before the Gemini step.
Nesting the iterator and array aggregator in an attempt to get the iterator to stop sending more operations, although that doesn’t work for my use case.
I also bent on investigating about the batch-processing capabilities of the Google Cloud Vision API, but I’m not sure if that works in Make.
I have tried to look in the Make forum for a workaround to break off an iterator mid-stream without success.
Here is my blueprint from my make scenario:
Note saving_part_1.blueprint.json (95.9 KB)
But I still don’t know:
How do I ensure that the Gemini module or any module runs only once, after the Array Aggregator has finished?
Is it possible to send a batch of data in Make to Google Cloud Vision using a single API call?
Are there more effective ways to extract OCR information from different file types like images and documents and then pass an aggregated single result to an AI module?