What is your goal?
Agent watches google drive folder, sees new screenshot → Downloads file from Google Drive → MakeAI content extractor uses an OCR on the image and gives back the fields in a bundle → Bundle information is mapped to a data structure or variables and injected into a google sheet template
What is the problem?
I need to remove 72 pieces of data and map them so I can put it into a Google Drive template.
What have you tried so far?
I have tried to use the JSON parse agent on the output from the Extractor and an Iterator as well (Super new to this)
Screenshots: scenario setup, module configuration, errors
Hey there,
what are you trying to extract exactly? Maybe you can use a text parser to find the strings you need if its something static.
The real fix is upstream. Don’t bother trying to parse 72 unlabeled values out of plain text after the fact, the OCR module itself doesn’t take a custom schema, this is why you are getting a flat list. How about feeding that raw text into an OpenAI “Transform Text to Structured Data” module right after the extractor. You can define your exact field names there “MATCH RANK”, “PLAYERS SAVED” and tell it to return valid JSON with those keys. And then a single Parse JSON module gives you named fields that you can map straight into your Google Sheets template, no iterator or regex needed. By doing this you can save a ton of module overhead. Another tip is if some fields are inconsistently detected, add “use null if not found” to the prompt so that your JSON structure stays stable across runs.
