← All work

Case study 07 · Marketing

AI Content Engine

Trending news → brand-voice caption → generated image → one-click approval → Instagram & LinkedIn.

  • n8n
  • GPT-4o
  • LangChain
  • Image generation API
  • Instagram API
  • LinkedIn API
per post, from research to published
Hours → 1 click
per post, from research to published

The problem

Researching, writing, designing and posting took hours per post, so the brand posted rarely and inconsistently.

What we built

  • Sources and ranks trending industry news daily
  • Drafts captions and hashtags in the brand's voice
  • Generates a matching image
  • Queues everything for human review — one click to approve
  • Schedules and publishes to Instagram and LinkedIn

How it flows

From trigger to done, step by step.

  1. Daily scheduleNews sources scanned
  2. Ranks topicsWhat matters today
  3. Writes captionBrand voice + hashtags
  4. Generates imageOn-brand visual
  5. 1-click approvalHuman stays in control
  6. PublishedInstagram + LinkedIn

What changed

  • ✓Consistent posting without a content team
  • ✓A human approves everything that goes out

Try it yourself

Infinityy OS runs the same kind of pipeline, live — with fictional data.

Live demo

See it inside Infinityy OS →

Your business

Build something like this →

Next case study

RegWatch →