Skip to main content
← Back to projects
✍️

LinkedIn Content Automation

Prototype

Taking the Content Orchestra Pipeline to its logical conclusion: a fully automated publishing system with human-in-the-loop approval.

The vision: specify a topic and target audience, the system researches (80+ sources), writes the article, generates images, drafts a LinkedIn companion post, and queues everything for one-click approval on the dashboard.

Currently the pipeline handles research-to-deployment. The next phase adds the approval dashboard, LinkedIn API posting, scheduling for optimal engagement times, and performance metrics tracking.

Tech Stack

Claude Sonnet 4Gemini ProLinkedIn APINext.jsVercel

Key Features

  • One-click article generation from topic
  • Automated research with source verification
  • AI image generation (Nano Banana Pro)
  • Companion post drafting for LinkedIn
  • Approval workflow dashboard
  • Optimal time scheduling (Tue-Thu 08:00-09:00 CET)
  • Performance metrics and engagement tracking

Challenges

LinkedIn's API doesn't support long-form articles — only regular posts with links. The companion post strategy (short hook + blog link) actually performs better algorithmically. The real challenge is maintaining authentic voice across AI-generated content.