I made coffee, moved the folder to the kitchen table, and opened it.
What I found was a small archaeology project.
Utility bills from an apartment I moved out of in 2021. Three years of car insurance correspondence, including renewal reminders for a policy I had cancelled. Receipts for furniture bought from a store that had since gone out of business. Bank statements from a joint account I had not thought about since a relationship ended. A lease agreement — unsigned — for a flat I had apparently decided against at the last minute.
And somewhere in the middle, mixed in with the detritus: things I had actually been looking for. The original purchase receipt for my laptop — useful for the insurance claim I had given up on months ago. A copy of a freelance contract with a clause I had disputed and never resolved. A medical reimbursement form, still unfilled, from 2023.
The problem was not chaos — it was density
What struck me was not that the documents were in disorder. Some were. But mostly they were just present. Densely, undiscriminatingly present. A lease renewal had the same physical weight as a notice from the tax authority. A coffee shop receipt occupied the same space as a dividend statement.
The problem with managing paper this way is not that you cannot find things — it is that you cannot quickly tell what matters. You have to read to triage, and you have to triage before you can read properly. It is a loop.
What I needed was a first pass. Something that could look at each page and say: this is a financial document, this is a contract, this is junk. Not perfectly — just directionally.
Scanning the pile
I had a decent phone scanner and enough patience to batch things into a few multi-page PDFs. That part took about two hours — less than I expected, more than I wanted.
Running them through PaperSweep took a few minutes more. The result was a spreadsheet with each document listed by category, alongside a confidence score and an OCR preview I could scan quickly to decide whether the label was right.
Most were right. The car insurance letters landed under "Insurance." The bank statements under "Financial." The lease under "Property." A few medical documents clustered correctly. The unsigned lease was flagged as a contract — also correct, though I might have labelled it "near miss, former life."
What the exercise actually revealed
I had been expecting a boring administrative outcome: documents filed, nothing surprising. Instead I found three things worth acting on.
The laptop receipt meant I could reopen the insurance claim. It was still within the extended window — I had not checked because I had assumed the receipt was gone.
The freelance contract dispute had a two-year clause I had forgotten about. Still enforceable, still within reach.
The medical reimbursement form was for €340. I filled it in the same afternoon.
None of this required a sophisticated system. It required knowing what I had. The document chaos was not producing entropy — it was producing invisibility. Things did not cease to exist when they disappeared into the folder; they just ceased to be actionable.
What I do differently now
The new rule is simple: anything on paper that arrives gets scanned the same week and uploaded in a batch. The physical originals go into a single, undivided folder labelled by year. I do not sort them further. PaperSweep produces the index; the folder is the archive.
The time cost is about fifteen minutes a month. The return — from that first session alone — was more than €400 and one resolved dispute.
The accordion folder is still under my desk. But now it only has one year in it.