Extract PO PDFs to Google Sheets + EDI 850 with n8n

Turn PO PDFs into structured Google Sheets—and optionally EDI 850—automatically with n8n

This n8n workflow collects one or more purchase order (PO) PDFs from an n8n Form, extracts PO header + line-item data with easybits Extractor, appends clean results into Google Sheets with duplicate detection, and can generate an EDI 850 file per new PO (saved to Google Drive).

What this workflow does

• Receives PO PDFs via n8n Form, including a toggle to enable/disable EDI 850 generation.
• Loads existing Google Sheets rows to build a list of already-processed PO numbers for deduplication.
• Splits and processes uploads sequentially into one PDF per item for reliable extraction.
• Extracts PO data using easybits Extractor, including PO header fields and the full articles table.
• Skips duplicates: if the extracted PO number already exists in Google Sheets, those documents are not re-added (duplicates are tracked for the summary).
• Flattens line items for reporting: writes one row per article line item to Google Sheets, calculating totals and flagging missing fields.
• Optionally generates EDI 850: reshapes extracted PO data into a canonical object, runs a sub-workflow to create an .edi file, and uploads it to Google Drive.
• Shows a completion summary with clean extractions, duplicates skipped, and any missing fields detected per document.

Use cases

• Centralize PO intake: upload supplier PO PDFs and keep procurement data synchronized in Google Sheets.
• Prevent duplicate processing across team submissions using PO-number deduplication.
• Generate EDI 850 documents only for newly received POs and store them in Google Drive.

Technical details

• Integrations/logic: easybits Extractor, Google Sheets (append + dedupe), Google Drive (upload .edi files)
• n8n nodes mentioned: form, form trigger, if, code, sticky note, google drive

Setup note: Create and connect an easybits Extractor API credential at https://easybits.tech/ and configure the extractor fields to return PO header values and the articles array.