Skip to main content

Example: Daily Excel Report

Scenario: You receive invoices daily and manually type data into an Excel spreadsheet. Let's automate this.

The manual process (before)

  1. Open the invoice PDF
  2. Read the invoice number, date, vendor, amount
  3. Open your Excel spreadsheet
  4. Type each value into the correct column
  5. Repeat for every invoice

Time: 5-15 minutes per invoice. With 20 invoices/day, that's 2-5 hours of manual data entry.

The automated process (after)

Step 1: Upload your Excel template

Upload the spreadsheet you normally fill. Make sure column headers are clear:

Invoice #DateVendorDescriptionAmountTaxTotal

Step 2: Create an extraction template

Create a template with fields matching your spreadsheet columns:

FieldTypeDescription
Invoice NumberTextThe invoice or reference number
Invoice DateDateDate the invoice was issued
Vendor NameTextCompany or person who issued the invoice
DescriptionTextBrief description of goods or services
Subtotal AmountNumberAmount before tax
Tax AmountNumberTax amount
Total AmountNumberFinal total including tax

Step 3: Batch process invoices

  1. Open the Document Parser
  2. Drag all your day's invoices into the upload zone
  3. Select your extraction template
  4. Click Start Parsing

All invoices are processed simultaneously.

Step 4: Fill the spreadsheet

  1. Open your Excel template
  2. Click Auto-Detect Mappings
  3. Watch as each column fills automatically
  4. Review and make any corrections

Step 5: Export

Click Export to download the filled spreadsheet.

Time: 2-5 minutes total, regardless of how many invoices.

Taking it further: Full automation

Turn this into a zero-touch workflow:

Schedule (daily 8 AM) → Check email inbox → Parse new invoices →
Fill Excel template → Send filled report via email

See Email-to-Data Pipeline for how to set up email-triggered processing.

Tips for this workflow

  • Test with 3-5 invoices first to verify template accuracy
  • Use specific column headers — "Invoice #" is better than "Number"
  • Add descriptions to the extraction template for any fields the AI gets wrong
  • Keep a consistent invoice format — if your vendors have wildly different layouts, you may need to review more carefully