Skip to content

Invoice Preprocessing

The PDF preprocessor automates invoice renaming using OCR, entity matching, and batch file saving. Upload invoices, review extracted data, and save directly to the REA Invoices folder on the server.

Processing Invoices Workflow

  1. Upload PDFs
    Drag multiple PDF files into the upload box or click Browse files.

  2. Click Process PDFs
    The system extracts data using OCR and matches against your databases. Give it time to load, processing multiple files takes a moment.

  3. Review Results
    Each PDF shows a preview and extracted details.

  4. Saving Results
    Check Save to REA Invoices and save the batch to the REA Invoices folder on the server.

Document Preprocessor dashboard

Updating the Vendor Database

In addition to the entity database, invoice processing uses a vendor database.

To update the vendor database:

  1. Click View / Edit Vendor Database dropdown
  2. The editor shows vendor names and matching patterns
  3. Click the + button (top right) to add a new entry
  4. Edit existing entries by clicking into the cells
  5. Click Save Vendor Database to save changes

Reviewing and Correcting Results

After processing, the dashboard displays the PDF preview on the left and extracted details on the right:

PDF preview with extracted details

Fields Explained

  • Due Date: Invoice due date extracted from the invoice
  • Date: Extracted in YYYYMMDD format (today’s date if invoice date is unclear)
  • Mortgage: Toggle for mortgage-related invoices (auto-detected)
  • Utility: Toggle for utility invoices (flags vendor as utility provider)
  • Address: Property address extracted from the invoice
  • Entity Property: Legal entity that owns or manages the property
  • Property Name: Building or tenant name
  • Vendor: Service provider or vendor name
  • Invoice Number: Invoice ID from the document

Mortgage Detection

The Mortgage toggle uses a two-step detection system:

  1. Address + Vendor combo check: The system maintains a database of known mortgage combinations. When an invoice matches a saved address/vendor pair, it flags as a potential mortgage.

  2. Keyword verification: The system scans the document for mortgage-related keywords (e.g., “mortgage”, “escrow”). Both checks must pass for auto-flagging.

When you manually enable the Mortgage toggle, the address/vendor combination saves to the mortgage database. Future invoices with that same combo will auto-flag as mortgage.

Utility Vendor Detection

The Utility toggle marks vendors as utility providers. When checked:

  • The vendor saves to the vendor database with a utility flag
  • Future invoices from that vendor auto-enable the Utility toggle

Using Dropdowns to Correct Data

Each field shows a dropdown with the best matches. The confidence score appears in parentheses. e.g. (0.55) means 55% confidence.

Due date matching:
Due dates are automatically extracted from the invoice using keywords. Click on the date box to change the extracted date. The calendar picker allows you to select the correct due date if the automatic extraction is incorrect.

Leave the field blank if the invoice has no due date.

Due date calendar picker

Address matching:
The most efficient approach is to search by address. Begin typing in the address dropdown. The system will auto-populate Entity Property and Property Name when you select a match.

Address dropdown with confidence scores

Invoice number matching:
If confidence is low, the invoice number field will be blank. If detected, it auto-selects the best match. If wrong, review other options in the dropdown:

Invoice number dropdown with alternatives

Invoice-Specific Manual Overrides

  • Address manual override: Clears entity property and property name, using only the manually entered address.
  • New vendor override: Manual vendor override automatically creates a new vendor entry in the vendor database.
  • Generated filename override: The text box under Generated Filename can also be edited directly. Any edits here take precedence over all other fields.

Manual override mode

Split PDF

Split PDF controls

Split multi-page PDFs into separate documents for individual processing. This is useful when multiple invoices are combined in a single PDF file.

How to split:

  1. Enter the number of pages per split (e.g., 2 to split every 2 pages)
  2. Click Split Current PDF
  3. The system creates new PDFs and processes each one automatically

Example: A 6-page PDF split every 2 pages creates:

  • PDF 1: Pages 1-2
  • PDF 2: Pages 3-4
  • PDF 3: Pages 5-6

Saving to REA Invoices

Invoices save to the REA Invoices folder on the server, organized into dated subfolders.

  1. Check Save to REA Invoices
    In the Save PDFs panel, tick the Save to REA Invoices checkbox.

  2. Click Save All to Server
    The system saves the entire processing stack to the REA Invoices folder on the server.

Save Location

Invoices route to the REA Invoices folder, with a dated subfolder in mm-dd-yyyy format:

Z:\Shared\BBT Asset Management Inc\Accounting\REA_Invoices\mm-dd-yyyy\