Built something cool this week — sharing the full breakdown here.

CLIENT PROBLEM:

A product brand was manually creating ad creatives for every launch. Designer, briefing, revisions, approval, export. 3-4 days minimum. $200-500 per project.

WHAT WE BUILT:

Full end-to-end AI Ad Creative Automation System using n8n.

Here is exactly how it works:

1. Client opens a form (no login, no tech knowledge needed)

2. Uploads ONE product image + fills 5 fields (product name, audience, benefits, platform, style)

3. AI (Google Gemini) reads the actual product image — analyzes packaging color, material, label, shape

4. Writes detailed image generation prompts for each variation

5. AI image model generates up to 4 ad creatives automatically

6. Client receives an email with image previews — clicks Approve or Reject

7. Approved images upload directly to Google Drive, organized by job

RESULTS:

• Time: 3-4 days → under 5 minutes

• Cost per launch: $200-500 → near zero

• Variations per run: up to 4

• Platforms covered: Instagram, TikTok, Facebook, YouTube (correct sizes auto-applied)

• Human involvement: zero (after form submission)

STACK:

n8n (workflow engine) + Google Gemini 1.5 Pro (vision + prompt writing) + Imagen AI (image generation) + Gmail (approval flow) + Google Drive (storage)

TOTAL NODES: 16

TOTAL SETUP TIME: 1 week including debugging

This is a real client project — not a concept.

Happy to answer any questions about how it works or how to build something similar.

2 comments