Gather documents
Invoices, receipts, quotes, delivery notes, contracts and purchase orders, stored by type. Each file is checked to be what it claims before it is taken in.
Transaction documents, kept where you can find them.
Gathers invoices, receipts and quotes and keeps them searchable by date, partner and amount — as a cloud service a team shares in the browser on Windows or Mac, or as a desktop app that keeps everything on one computer. It is in preparation.
Terms and policies
The terms of use and privacy policy will be linked here at release.
All terms and policiesInvoices arrive as attachments, receipts as downloads, quotes in a shared folder. When they come in different ways, finding the paperwork for one transaction takes longer every month. APOUL E-File gathers them in one place and files each with its date, partner and amount.
Invoices, receipts, quotes, delivery notes, contracts and purchase orders, stored by type. Each file is checked to be what it claims before it is taken in.
AI reads the date, partner, amount and registration number and offers them as suggestions, alongside checks such as whether subtotal plus tax matches the total.
A document only becomes confirmed after a person has checked its date, partner, amount and type.
Find every document for a partner or a period without opening folder after folder.
Owner, admin, editor and viewer roles decide who sees and changes what.
Originals cannot be changed afterwards, and every action is logged.
You should be able to guess the next step without reading a manual.
Add the document files from the browser.
Compare what AI read, and correct it where needed.
Confirmed documents can be filtered by date, partner and amount, and the list exported as CSV.
The original file and its hash cannot be changed after it is stored; deleting a document is handled as a request. The audit log and the history of document details are append-only, and each log entry is chained to the one before it so tampering shows. Organisations are kept apart, access is split into four roles, joining is by invitation only, multi-factor sign-in with an authenticator app is required, and access can be limited to set IP addresses. The service is designed to run in a data centre in Tokyo. Whether it meets the requirements of Japan’s Electronic Books Maintenance Act has not been confirmed yet, and will be stated here once it has.
The same product in two forms, depending on how many people use it and where the documents should live.
Upload documents in the browser; a person confirms the date, partner and amount that AI suggests. Owner, admin, editor and viewer roles decide who sees what, and every action is logged. Works on Windows and Mac.
Drag a document onto the window and it reads the transaction date, partner and amount, then saves it under a name such as “20260731_SampleTrading_203600.pdf” in folders by year and type. Reading happens on the computer; documents are not sent anywhere. It runs on Mac today, and a Windows version is being prepared.
The desktop version takes documents by drag and drop, from a watched folder (which can be a scanner’s save folder), or straight from email.
The desktop version takes PDF, images (JPEG, PNG, HEIC and more), email (.eml, .msg), ZIP, Word and Excel.
Japanese era dates and amounts written in kanji are read, and a payment due date is not mistaken for the transaction date.
The problems this app is built around, and what to settle before you bring it in.