Get a demo

Freight invoice digitization

Extract every charge from every freight invoice PDF. Classify freight, handling, and customs automatically. Stop the per-invoice 30-minute parse.

Get a demo

The prompt

I want to extract and categorize charges from a freight invoice PDF. Build me a flow that pulls each charge line, flags anything past due based on today's date, and categorizes each charge by freight, handling, or customs.

Just copy and paste the prompt into a new Parabola flow to get started.

What Parabola builds

A workflow with five steps you can edit:

1. Receive the invoice PDF. Forwarded from forwarder email, uploaded to a folder, or pulled from a billing portal.

2. Extract the charge lines. AI parses every line into charge code, description, amount, and date.

3. Categorize each charge. Freight, handling, customs, fuel surcharge, accessorial. The flow maps the forwarder's wording to your internal taxonomy.

4. Flag past-due lines. Compare each charge date against today. Anything outside your payment window gets surfaced.

5. Output the table. Charge-level detail plus a category summary, dropped into a Parabola Table or sent to AP.

Why teams stop doing this manually

Freight invoices are the worst kind of PDF. Twenty pages long. Charge lines split across multiple sections. The same charge type called "freight" by one forwarder, "transport" by another, and "carriage" by a third. AP needs to categorize each line correctly to post the right GL entry. They cannot do that until somebody reads the PDF.

The manual version is an analyst who opens each invoice in Adobe, types the charges into a spreadsheet, classifies them by category, and uploads the spreadsheet to AP. Thirty to sixty minutes per invoice. Two to three hundred invoices a month at a mid-sized brand. The math does not work.

The shortcut is to post everything to a single "freight" GL bucket and call it good. That gets the invoice paid but ruins category-level cost analysis. Finance ends up asking the same question every quarter: "Why do we not know what we actually spend on customs?"

How it works

Step 1. Paste the prompt.

Open Parabola, paste the prompt in section 2, and let it ask follow-up questions about your forwarders, charge taxonomy, and AP workflow.

Step 2. Connect your data.

Invoice inbox or billing portal feed, your internal charge category taxonomy, AP destination.

Step 3. Run it on every invoice.

Triggered per email or run on a batch. The output goes to AP, GL, or wherever the categorization lives.

Related prompts

FAQ

Does this work for PDFs from multiple forwarders with different layouts?

Yes. The AI extraction handles layout variance. The classification step maps each forwarder's charge codes to your internal taxonomy.

What about scanned invoices or low-quality PDFs?

The AI step handles scanned PDFs and OCR-required documents. Quality varies, so the flow surfaces low-confidence extractions for human review before posting.

Can I push the output directly to NetSuite or SAP?

Yes. The flow outputs a journal-entry-ready table. Some teams write back to the ERP via API, others output a CSV for the AP team to import.

How does the past-due flag work?

Each charge line has a date. The flow compares against today and your payment-term rule. Charges past their payment window get flagged in the exception view.

What if the forwarder changes their charge code names?

Update the mapping row. The classification step pulls from the mapping table, so adding a new code or renaming an old one is a config change, not a workflow change.
Stop reading PDFs to post a journal entry.
Paste the prompt, point it at your invoice inbox, and let the parsing run on its own.
Start for free