AutoIndex in a file cabinet against the same file cabinet-1539982831

Can anyone clarify this for me? I'm running an AutoIndex job where the file cabinet is also the external data source. I'm doing this to weed out duplicate invoices. I match on invoice number, amount and vendor. On the first pass I filter for status 'New Invoice' and check against any invoices that are NOT status 'New Invoice'. The AI job runs every 10 seconds. The end user uses an Ad Hoc workflow process using Task Lists and Stamps that change the Status field (Version 6.6) to 'Processed'.

The problem is of course that if 2 users store the same invoice before one of them gets processed, the duplicate is not found because they are both 'New'.

My question is, if I change the filter and just check against all invoices regardless of Status, will the AI job 'find itself' so to speak and mark all "New Invoices" as duplicate? Or does AI not look for the same docID in the comparison set as it has currently in memory?

My head hurts,

Hmmm. Keep in mind that the external source and the cabinet are 2 separate entities, even though they are one and the same. Your filter should include anything not marked as a duplicate. It is not looking at the docid at all. It is looking at 2 records that match. Therefore if you do not filter on the status it will probably call the same record a duplicate.
Other than that I don't have a solution right now.

Phil