The problem
The client is a B2B custom-packaging company. Every customer request moved through several disconnected tools: email, spreadsheets, supplier documents and a legacy ERP.
The same numbers were typed by hand into more than one system as a request went from quote to order to invoice. Each hand-off was manual, slow, and a chance for errors.
What was built
An internal ERP covering the full order lifecycle, from first request to invoice, with n8n running the automation underneath it.
- Incoming requests are captured as structured records instead of being re-typed.
- AI extracts the relevant data from supplier documents, with a human review step for anything it isn't sure about.
- Invoicing and payment tracking are connected to the accounting and banking systems, so they happen as part of the workflow instead of as separate data entry.
How the n8n layer is maintained
Workflows are treated like application code: changes are version-controlled, reviewed and tested before they reach production, not edited directly on the live instance.
Stack
- Orchestration: self-hosted n8n
- Application: custom web app on a Postgres database
- AI: large language models for document understanding
- Integrations: accounting, banking and logistics providers
Credentials
Our Apollo.io node is published on npm as
n8n-nodes-apollo (MIT, maintainer arturl95,
14 resources, source public). We hold n8n Verified Creator status.