Get a feel of the flow of SimpleIndex

Drives your scanner through its TWAIN driver — duplex, ADF, blank page removal, deskew, and the rest of the hardware's own settings. Nearly every desktop and production scanner made in the last twenty years ships with one.

The other driver standard, common on high-volume production hardware and often faster than the same machine's TWAIN driver. If you bought the scanner to run all day, this is the one to use.

Point any device at a folder and SimpleIndex processes whatever lands there. This is how network copiers and MFPs work without a driver, and how you replace a per-click capture license.

Pull attachments straight out of a monitored mailbox, and convert the message body itself to PDF when the email is the document. Common for invoices and remittances that arrive by email.

When a file already contains real text, read it directly instead of running OCR on a picture of it. Faster and exact — no recognition error, because nothing is being recognized.

Read values straight out of structured files rather than off a page image, and write them back out the same way. Useful where a system already exports data but the document has to be filed alongside it.

A phone camera as the scanner for field staff and branch offices. Pages captured to a shared cloud folder get picked up and processed like anything else.

Read text off a region of the screen — useful for pulling values out of an application that has no export, no API, and no printing worth using.

Read one value off the page - an account number, a claim ID - and look up everything else from your existing records. Limiting the recognized fields on the page limits the chances data will be read and stored incorrectly

A value that doesn't match a record in your database gets flagged rather than filed. This catches the OCR error that would otherwise become an untraceable document six months from now.

Turn on a review step when you build the job and every recognized field is shown for a person to confirm or correct before anything is written. Turn it off and the batch runs straight through — which is the right call for barcodes and the wrong one for a first pass at OCR.

Constrain a field to what it can legitimately be — a date, a nine-digit number, a value from a list. A malformed read fails the rule instead of being written into a filename.

Decide where one document ends and the next begins — by barcode coversheet, patch code, page count, or a change in the value read off the page. Scan the whole box at once and let the batch split itself.

Work out what each document is from its content, then apply the right rules to it. One mixed stack of applications, IDs, and disclosures sorts itself into three named sets.

Define what a complete file looks like and get told what's missing as the batch processes, not at audit time. A loan file with no disclosure comes back as a list rather than a finding.

Split two-up scans into single pages, drop blanks, deskew, crop, and despeckle before recognition runs. Better images mean better OCR, and this is where most accuracy problems actually get solved.

Build filenames and folder paths from any combination of index values, dates, and fixed text. Your naming convention becomes a rule the software follows instead of a document nobody reads.

The simplest output and often the right one: named PDFs in a structured folder tree, each with a searchable text layer. No platform to buy, no vendor to depend on, readable in twenty years.

Write index values and file paths into your own tables so documents are queryable alongside the rest of your data. Also exports to CSV and XML for systems that prefer a file.

Upload into libraries with index values mapped to SharePoint columns, so documents arrive filterable and searchable rather than as a folder of PDFs somebody has to sort out later.

For applications with no import path at all, robotic process automation drives the interface the way a person would - open the record, type the fields, attach the file, save. SimpleIndex doesn't contain the robot; it starts one, from the command line or the Workflow Manager, and hands it the values it read.

Stamp page numbers, Bates sequences, dates, or watermarks onto the output. Standard requirement for legal discovery and for anything that has to be produced in a numbered set.

The reverse direction - take structured data and produce completed PDF forms in bulk, for filings that have to go out on a specific official form.

Email the finished document out as it is filed, to an address that can be built from the values read off the page. Runs over SMTP, or through the Office 365 and Gmail APIs where a modern tenant requires OAuth rather than a password.

POST, PUT or PATCH the finished document to any HTTP endpoint, as multipart or a raw body, with the job's index values sent alongside the file so the receiving service can identify it. Authenticates with a bearer token or a username and password, sends custom request headers, and retries automatically on the failures where a retry helps. The response can be captured back into an index field, either whole or as a single value pulled from JSON.

SimpleExport converts index data and files into whatever layout a content management platform expects, for the systems that want neither a database row nor a folder. Included with Professional from v12, and still sold on its own.

SimpleQB posts captured invoice and receipt data into QuickBooks, matching vendors and accounts as it goes. Included with Professional from v12 — and still available on its own, because plenty of sites want Standard plus QuickBooks and nothing else.

Hide the entire configuration behind a single button for the person who runs the batch. Pre-configure jobs that are ready to use by any staff, from novice to senior executive.

Every setup saves as a job file you can copy, version, and move between machines. Build the invoice job once and deploy it to four workstations, or hand it to a branch office.

Chain several jobs into one sequence: capture a batch, export the data, print coversheets, deliver the files. Each step names the job it runs and can open that job for editing, and the log appears while the step is still working. A step that fails stops the run rather than letting the rest carry on.

Runs as a Windows service with no operator and no login session. Batches dropped in a watched folder at 6pm are processed and filed by morning, on a schedule you set.

Trigger jobs from a script, a scheduler, or another application, and pass parameters in. The hook for anyone who wants SimpleIndex to be a step inside a larger automated process.