Hi,

I want to ask for Ideas to optimize AI Attachment Analyze workflow
My files are named with random group of chars.
I created a table, where Attachment uploaded to ‘File’ field. There is AI field READ that analyze content of Attachment, and outputs a few parameters needed to categorize.
For Example, it’s ‘Doc Number’, ‘pages’, ‘City’ ‘Doc date’, ‘Summary’. City is single-select. 
After AI filled this data, automation triggered, that puts each data in respective field
Then formula field ‘New filename’ generates name according to this data.
Then automation launched by ‘New filename’ renames a file and puts it to ‘Document’ field
(i suppose that each record has 1 file only, if more, it gets only with first in cell. it’s OK)

it all works OK, but since I don’t need 2 doc fields, I changed Automation so it renames file and puts back into the same ‘File’ field . But this triggers AI again.  AI gets the same data so nothing changes after that. But I don’t like that it makes second run. It’s not just a double waste of AI credits (yes, I chose less expensive AI model capable of doing this task), but also slower total task execution, when I put for example 40 files at a time.  
Even if I uncheck ‘Automatic run’, when I upload new files, AI don’t run, there’s just a button ‘Run agent’.
But when I run agent, it does all that thing, filename changed, file saved back to the same cell and then AI reacts on file change (despite it’s only name changed) and runs ‘Analyze’ again

So, it started:

and that part I would like to avoid (analyzing back when I rename file)
 

The only way I see is to create some dummy field, filled by default, make it required for AI together with File, and then empty it before I rename a file. I din’t check it yet, but I already don’t like it. Also I prefer to not use 3rd party tools if the task is doable with Airtable only. 
Is there any better way to do it?