Built a production-grade AI content pipeline in n8n — looking for feedback
Hey everyone, Abdullah here 👋 I’ve been building production-grade n8n systems and recently finished an agency-level AI content pipeline. Thought this community might appreciate a real backend example instead of a toy demo. What the system does: • Frontend form → structured campaign request • Backend n8n workflows handle generation, validation, and state tracking • Human-in-the-loop approvals directly inside Slack (Approve / Reject buttons) • Auto-scheduling after approval • Simulated analytics ingestion (likes, comments, etc.) stored in DB • Live status updates so teams don’t have to chase each other Key focus areas were: • Reliability (idempotency, state handling) • No “prompt spaghetti” • No forcing users into new dashboards • Designed to plug into real teams (Slack, Notion, CRM, etc.) I recorded a demo video walking through the frontend → n8n logic → Slack approvals → scheduling flow. 🎥 Demo video: https://www.youtube.com/watch?v=li3SiHrqIyM Appreciate any feedback 🙏