Help for owners
Documents pile up with random names and nobody can find anything
Short answer
Document automation reads a file, decides what it is, renames it by a fixed pattern and puts it in the right folder. It fails on poor scans and unusual layouts, on name clashes and duplicates, and when the folder rules do not match who is allowed to see the document.
Does this sound like your situation?
- Files are called scan0032.pdf and final_v2_real.docx.
- I search for an invoice for twenty minutes.
- Staff save documents wherever they like.
- The same contract exists in four folders.
- Someone could open a file that should be private.
Why does this happen?
There is no naming standard
Without a pattern such as date, party and type, names depend on the person. Searching by name then stops working.
Reading the content is unreliable
Software that identifies the supplier, date or amount from a scan can fail on blurry pages, skewed images, stamps and other languages. Treat what it reads as a suggestion.
Documents of one kind look different
Invoices from two suppliers have different layouts. A rule built for one fails on the next.
Duplicates and clashes
Two files with the same generated name can overwrite each other, or a document received twice can be filed twice.
Folders carry access rights
Moving a file changes who can open it. A payroll document filed in a general folder becomes visible to everyone who uses that folder.
What can I do about it today?
- Choose one naming patternFor example date, who, what and a number. Write an example and share it.
- Draw a simple folder treeA few levels at most, by client or by year. Deep trees are not kept up.
- Decide who may see whatNote which folders are private, such as people and finance.
- Do not rename the originals in bulk yetCopy the files first. Bulk renaming of the only copy cannot easily be undone.
- Find out which documents are legally keptRetention rules depend on your country, so ask your accountant or adviser before deleting anything.
When should I bring someone in?
- Thousands of existing files need sorting.
- Documents arrive by email or scanner and must be filed automatically.
- Files contain personal or financial data.
How do I handle it?
- Sample the real documentsI look at the types, languages and scan quality you actually have before choosing an approach.
- Work on copiesThe original folder stays untouched until the results have been checked.
- Combine content rules with a review queueDocuments that are read with confidence are filed. Doubtful ones go to a list for a person to confirm.
- Prevent clashesA unique number is part of each name and duplicates are detected before filing.
- Respect accessFolder permissions are checked, so a filed document ends up visible only to the right people.
The price depends on what I find, so I quote in writing after a short discovery call. Tell me what is happening and I reply within 24 hours.
What do owners ask about this?
Can it sort old files too?
Yes, on a copy. Quality depends on the scans and a share of files will still need a person.
Where should the files be stored?
In the storage you already use, such as a shared drive or cloud folder. I do not recommend moving everything for the sake of the automation.
Is it safe to send documents to an AI service?
It depends on the content and on that service's terms. For sensitive documents I prefer processing that stays within your own accounts, and I will tell you what leaves them.
How was this page made?
I drafted this page with AI assistance. It is general guidance for business owners, not advice for your own system. Found a mistake? Tell me.