Structured accounting documents: from clutter to actionable file

Before accounting, document clutter must become actionable documents — separated, identified, renamed and ready for your file.

The document clutter problem

Many accounting files start as a mix: multi-page PDFs, poorly named scans, invoices and receipts in the same file. Reading a document is not enough — you must classify, rename and make it identifiable for the rest of the accounting work. Concretely: documents mixed in a single PDF, files named Scan001.pdf or Final_v3.pdf, no way to find a specific document without manual sorting, files hard to send to a fiduciary, time lost separating and renaming before entry. Example: in the Alpine Menuiserie SA demo scenario, a batch of cluttered documents becomes nine distinct purchase documents, renamed and numbered — a 10-page mixed PDF becomes 9 separate supplier invoices and 1 bank statement, each renamed by date, supplier and amount and numbered 01_, 02_… — ready for the rest of the file.

What Cortex does to structure your documents

1. Drop the clutter — Drag and drop multi-page PDFs, scan batches or photos. One upload is enough — Cortex accepts clutter as-is.
2. Document separation — Each supporting document is isolated: 1 invoice = 1 document. Multi-page PDFs are split into distinct files.
3. Identification — Date, supplier, amount and document type are identified to prepare renaming and numbering.
4. Consistent naming — Files are renamed by date, supplier and amount — readable at a glance, actionable per mandate.
5. Visual check — You verify structured documents before moving to entries. Structuring stays under your control.

What Cortex does not claim to do

• Not perfect OCR — Cortex reads and structures — but human review remains necessary on ambiguous cases.
• Not filing without review — Structuring reduces manual sorting; it does not promise filing free of any verification.
• Not zero entry — Structuring documents is step one. Entries are proposed next — and validated by you.
• No machine-only acceptance — Cortex prepares documents. Validation of entries and the final file remains in your hands.

Use cases

• SME — Reduce sorting and manual re-entry on document clutter while keeping entry validation.
• Fiduciary — Receive structured client files (numbered documents, proposed entries) without changing ERP.
• Freelancer — Upload supporting documents, validate proposals, export to your usual software.

Swiss practical cases — structuring clutter

Demo scenario: Joinery SME (Vaud) — mixed PDF batch

Demo scenario: Alpine Menuiserie SA uploads a 10-page PDF: Swisscom, Aligro, hardware invoices and a statement. Cortex isolates 9 purchase documents + 1 statement, renamed 01_2026-01-15_Swisscom_….pdf… You verify separation before any entry.

Restaurant (Geneva) — mixed tickets and invoices

One scan mixes till tickets, beverage invoices (8.1%) and food (2.6%). Structuring separates each document so VAT split can be proposed later — without validating on your behalf.

Fiduciary firm — client clutter per mandate

The client sends a messy Drive folder. Cortex structures per mandate: numbered, readable documents. The fiduciary receives a usable batch, not a bag of Scan001.pdf files.

Unstructured clutter vs structured documents

Naming convention — before / after

Where to go by intent

Responsibility for structuring

• Swiss document context — QR invoices, multi-rate VAT and fiduciary client batches: structuring prepares the local file, without promising perfect OCR or filing without review.
• You control the outcome — After separation and renaming, a visual check remains necessary. Cortex organizes clutter — you validate what follows (entries, export).
• Archive outside Cortex — Source PDFs are purged on the server within ~48h. Keep the ZIP on your side or at your fiduciary per your obligations.

FAQ — structured accounting documents

From clutter to actionable documents

See structuring demo · See pricing