Build n8n workflow for static Facebook ads using product images and research

I'm looking for someone who can help me build a local n8n automation flow for generating high-quality static Facebook ads based on product images and marketing research.

Here’s the general idea:

I upload:
- Product images (white background)
- A PDF file with market research (includes buyer profiles, pain points, buying motives, angles, etc.)

The automation should:
1. Read and extract key insights from the PDF
2. Determine the right ad strategy (based on Eugene Schwartz levels, market sophistication, emotional/rational buyer)
3. Choose the most suitable ad format (e.g. pain-point, review-style, before/after, etc.)
4. Generate 3–5 ad headlines to be placed on the image
5. Use AI image generation to place the product image into a realistic scene that fits the ad angle (e.g. product being used)
6. Add the headline onto the image
7. Output 4 ad variations (image + copy)

I want to be involved at key decision points (e.g. confirming angles, headlines, etc.) before it moves forward.

I run n8n locally and would like this to be fully local if possible.

Can you build this or something close to it?
Let me know if you need more details or if you can give an estimate (price + timeline).

Thanks!

Budget around 75$ for the whole flow

Please send me previous work + rate