n8n Invoice Data Extraction Workflow with Temlavo

Extract structured invoice data with Temlavo—automatically, reliably, and with safe recovery

This n8n Invoice Data Extraction Workflow uploads a single invoice file to Temlavo, polls until extraction is finished (or requires review), and returns a clean structured JSON result to your calling workflow—plus a recovery “safe-stop” when needed.

What this workflow does

• Receives one invoice item from a parent n8n workflow, including binary.invoice and a temlavo_submission payload with saved idempotency tokens and file metadata.
• Validates inputs by checking file type and size, and ensuring required Temlavo submission fields are present.
• Creates a Temlavo document via REST API using a fixed API request context and an Idempotency-Key header to prevent duplicate submissions.
• Uploads invoice bytes to Temlavo using the API-provided signed URL and headers, then confirms upload completion.
• Polls document status until it reaches a terminal state: completed, needs_review, or failed, using bounded waits and retry logic for transient API errors.
• Returns one JSON result to the caller with the final document payload (including review reasons when applicable), or a temlavo.n8n-safe-stop.v1 object describing how to recover.

Use cases

• Automatic accounts payable ingestion: extract invoice fields into structured JSON for downstream processing.
• Back-office workflows that need needs_review handling for ambiguous documents.
• Multi-tenant SaaS pipelines where reliable extraction and idempotency are critical.
• Engineers building end-to-end Temlavo-based invoice processing with controlled polling and safe-stop recovery.

Technical details

• Integrations: Temlavo via REST API (Idempotency-Key + signed upload URL) and polling for terminal status.
• n8n nodes/parts: code, wait, switch, http request, execute workflow trigger, and sticky note.
• Setup: Create a Temlavo API key (copy the full key once) and configure the HTTP request node accordingly.