How does InstantDrive name files automatically?
InstantDrive builds each name from the document type, the entity name, the identifier and a date, for example TradeLicense_Al_Noor_Trading_LLC_CN1234567_Exp20271231.pdf. When no expiry date is found it falls back to an issue date, then an amount, then the upload date.
What does automatic file naming do?
Phone photos arrive as IMG_8291.jpg and scanners produce Scan_0012.pdf. Names like these tell nobody what the file is, and a shared folder full of them cannot be searched or sorted. InstantDrive replaces the original name with one built from the content of the document after the document has been read.
The name is generated once, at upload, from the fields the reading step found. The same rule runs for uploads in the web app, photos from the phone, attachments that arrive at the company inboxes and files saved to the WebDAV drive. This page describes the pattern as the code produces it today, including the parts that differ from the pattern shown elsewhere on the site.
What is the real name pattern?
The generated name has up to four parts joined by underscores: a type prefix, the entity name, the identifier and a date token. If a document has no identifier, that part is left out and the name has three parts. The extension follows at the end.
The whole name before the extension is cut at 150 characters. Repeated underscores are collapsed into one.
| Part | Source | Example |
|---|---|---|
| Type prefix | A fixed label for the detected document type | TaxInvoice |
| Entity name | The company, person or supplier read from the document, cleaned | Gulf_Fresh_Supplies_LLC |
| Identifier | Licence, tax registration, ID, IBAN, Ejari or invoice number, with punctuation removed | 100234567890003 |
| Date token | Expiry date as Exp plus YYYYMMDD, or a fallback | Exp20261130 |
Which prefix does each document type get?
The prefix comes from a fixed list in the product. The pattern engine detects the types in the first table below. The prefix list also holds labels for further types such as purchase orders and SaaS invoices, which are used when a document has been given one of those categories. A category that is not on the list gets the prefix Document.
The vision fallback can return a few categories that have no entry on the list, such as a vehicle registration or a salary certificate. Those files also get the prefix Document.
| Detected type | Prefix in the name |
|---|---|
| Emirates ID | EmiratesID |
| Passport | Passport |
| Residence visa | ResidenceVisa |
| Tax invoice or receipt | TaxInvoice |
| Ejari or tenancy contract | Ejari |
| Bank statement | BankStatement |
| MOHRE labour contract | MOHREContract |
| Trade licence | TradeLicense |
| Commercial contract | Contract |
| Not recognised | Document |
How are the entity name and identifier cleaned?
The entity name is the company, person or supplier the reading step found. Everything that is not a letter or a digit becomes an underscore, so Gulf Fresh Supplies LLC becomes Gulf_Fresh_Supplies_LLC. Camera prefixes such as IMG, DSC, Scan, Photo and WhatsApp Image are stripped, as are long hexadecimal identifiers and a trailing file extension. Names that begin with Etisalat are shortened to Etisalat. If nothing usable is left, the entity becomes General.
The identifier keeps only letters and digits, so 784-1990-1234567-6 becomes 784199012345676 and CN-1234567 becomes CN1234567. Values such as NONE, NULL and N/A are treated as no identifier. Arabic text in an entity name is replaced by underscores, because only Latin letters and digits are kept. An entity name written in Arabic script alone therefore reduces to General.
What goes in the date token when there is no expiry date?
The last part of the name is always present, so every name ends with something sortable. The engine tries these sources in order and uses the first one that exists.
Three details matter. The amount is cut to a whole number, so 250.75 becomes 250. On a document with two or more dates, the reading step treats the last date as the expiry unless an expiry label is present. On an invoice, a due date is also treated as the expiry, so a supplier invoice often carries Exp followed by its due date. Check that token before you rely on it for tracking.
| Order | Source | Token | Example |
|---|---|---|---|
| 1 | Expiry date found in the document | Exp plus YYYYMMDD | Exp20271231 |
| 2 | Issue date, when no expiry date exists | YYYYMMDD with no Exp | 20260301 |
| 3 | Total amount, when the document has no date at all | Currency code plus whole amount | AED250 |
| 4 | Nothing found | The date of the upload | 20261106 |
What do real names look like for a fictional company?
Al Noor Trading LLC is a fictional Dubai company. The table shows names that the rules above produce for its documents. They assume the reading step found the stated entity, identifier and dates. If it reads a field differently, the name changes with it.
Two files that would receive the same name in the same folder are both kept. The second gets a number before the extension, such as _1.
| Document | Resulting name |
|---|---|
| Trade licence with an expiry date | TradeLicense_Al_Noor_Trading_LLC_CN1234567_Exp20271231.pdf |
| Employee Emirates ID photographed and uploaded as a .jpg | EmiratesID_Layla_Hassan_784199012345676_Exp20281103.jpg |
| Supplier tax invoice with a due date of 30 November 2026 | TaxInvoice_Gulf_Fresh_Supplies_LLC_100234567890003_Exp20261130.pdf |
| Ejari contract with one past issue date, no expiry found | Ejari_Al_Noor_Trading_LLC_20261234567_20260301.pdf |
| Cafe receipt with no date and no identifier, total AED 250 | TaxInvoice_Corner_Cafe_AED250.jpg |
| Unrecognised scan, nothing readable, uploaded on 6 November 2026 | Document_General_20261106.pdf |
Why do consistent names make search and sorting reliable?
File names are one of the four fields the vault search looks at, together with the entity name, the identifier and the category. When every name starts with a type prefix, typing TradeLicense or Ejari finds that kind of document. Typing the last digits of a licence or tax registration number finds the name that contains it. Typing a supplier name finds that supplier across invoices and statements.
Sorting by name then groups files by type first and by entity second. Within one entity and identifier, the date token is YYYYMMDD, which sorts in calendar order as plain text, so renewals of the same licence line up from oldest to newest. A date written as 01-12-2026 or Dec 2026 would not sort correctly. This is why the date format is fixed and has no separators.
How does the real output differ from the target convention?
InstantDrive documents its goal as DocCategory_EntityName_Identifier_ValidityDate.pdf. The generated names follow that shape closely but not exactly. The table lists the differences so that you can plan around them.
The extension point needs one more note. Photos are converted to WebP when that makes them smaller, but the name keeps the extension of the original upload. A name ending in .jpg can therefore point to WebP data. Do not rely on the extension to tell the format.
| Point | Target convention | What the product produces |
|---|---|---|
| Validity part | A validity date | Exp plus a date when an expiry exists, otherwise an issue date, an amount or the upload date |
| Extension | Always .pdf | The extension of the uploaded file, such as .jpg or .png |
| Spaces in entity names | Not specified | Replaced with underscores, so one name has many underscores |
| Missing identifier | A fixed slot | The slot is left out and the name has fewer parts |
| Unknown type | Not specified | Prefix Document, entity General when nothing is readable |
Does correcting a name teach the system new rules?
Not in a way you can depend on today. In the document panel you can correct the category, entity, identifier, dates and amounts of a document. That changes the stored fields of that one document. It does not rename the file, so a wrong name stays until the file is replaced or renamed through the mapped drive.
The product does contain a rule-learning component. After three identical corrections for the same tax registration number or supplier name it can create a routing rule that sets the folder, entity name and category of later uploads. It never changes the generated file name, and the correction screens in the app do not feed it in a way that reaches those rules. This page therefore does not describe learning as a feature.
What does automatic naming not do?
It does not guarantee that the name is right. The entity, identifier and dates come from reading software and can be wrong, particularly on poor photos. It does not rename existing files in bulk and it has no setting to change the pattern, the prefixes or the separators. It does not check a number against a government register. It does not turn Arabic company names into Latin transliteration.
Names are also not an access control or a security feature. They describe the file and nothing more.
What should you do after the first uploads?
Upload ten typical documents and read the generated names against the originals. Look at the entity, the identifier and the date token on each. Pay attention to invoices, where a due date can appear as an expiry, and to photos, where the extension may not match the format. Correct wrong fields in the document panel so that search and the expiry tools use the right values. The free file name generator on this site shows the same kind of pattern without an account.
Keep these documents in one drive
InstantDrive stores your licences, Emirates IDs, tenancy contracts and invoices, reads the numbers and dates, and names every file in one pattern.
Open the appPricing