AI-Driven n8n Workflow Automation

I’m building out a series of n8n workflows that will lean on AI to take repetitive back-office work off our hands, and I’m looking for someone who can own the technical build long-term.

First priority is a flow that logs in to Amazon and UPS seller/portal accounts, pulls the latest invoices, and stores them in a structured way. Where invoices arrive only by email, the same workflow should monitor a dedicated mailbox, pick up new messages, and use an extraction model to capture totals, dates, taxes, and line items with high accuracy.

The second piece is a cargo-shipment routine. When our team flags a new shipment, the workflow should read the job details and auto-fill the relevant online form—no manual re-keying. For now, it’s purely the form-filling step; tracking updates and confirmation emails can wait for a later phase.

I’m already using self-hosted n8n with Postgres, so nodes, credentials, and environment are ready. What I need from you:

• Robust n8n workflows (JSON) that handle login, scraping/API pulls, and email parsing
• Integrated AI steps for invoice data extraction, form completion, and smart error handling
• Clear, inline documentation so I can tweak field mapping without chasing variables
• A short Loom or similar walk-through of each finished flow

If this first sprint runs smoothly I’d like to keep iterating together, adding shipment tracking, confirmation messaging, and other back-office automations over time.