Skip to main content

Build Ai Content Pipeline With N8n

Xander Sebastian Xander Sebastian Published

How to wire together Claude, Perplexity, and your CMS in n8n to produce research-backed drafts on autopilot — including the exact workflow JSON.

1 min read

The content production pipeline we run for Robotic Pixels' own content and across client retainers uses n8n as the orchestration layer — Perplexity for research, Claude for outlining and drafting, the client CMS for output. Keyword to review-ready draft, without human intervention in the production stage. What follows is the architecture of that pipeline: the nodes, the API connections, and the workflow JSON. This is the current production spec, documented for audit and replication inside retainer setups where we build and operate the same infrastructure. It is not a starter template. It is what is running now.

Architecture overview

Trigger: manual or scheduled. Research node: Perplexity API call with your keyword plus competitor context. Outline node: Claude with a structured prompt that enforces your content framework. Draft node: Claude with the outline plus source material as context. Validation node: word count, heading structure, CTA presence. Output node: CMS API call to create draft post.

Continue reading

Ready to put AI to work in your marketing?

Book a Fit Call — 20 minutes to find out if we're the right fit. No pitch deck, no fluff. If we are, a Foundation Sprint sets the scope.