Create a fully automated system that generates complete product cards on Prom.ua in the status of draft based on data from Amazon ASIN, Google Sheets, OpenAI, and Prom API.
The system should automatically create title, SEO, description, specifications, keywords, photos, price, as well as generate content in Ukrainian and Russian.
Prohibited: any emojis, icons, decorative symbols.
1. Input Data (Google Sheets)
The table will contain:
The system should automatically process new rows.
2. Data Retrieval from Amazon
The system should retrieve:
Product title
Bullet points
Description
Technical specifications
Images (all available)
In case of Amazon blocking, a fallback method should be in place.
3. Content Generation via OpenAI
The system should generate:
3.1. Product Title
3.2. SEO Description (first 150 characters)
Embedded keywords
In Ukrainian and Russian
3.3. Full Description
Structured in blocks
H2 headings
Introduction about the product
Advantages
Technical specifications
Care recommendations
Block about product originality
Separate block about product condition (bold + italic)
In Ukrainian and Russian
3.4. Prohibitions
4. Image Processing
The system should:
Download photos from Amazon
Optimize (without loss of quality)
Rename according to SEO logic
Generate ALT texts (UA + RU)
Upload to Prom via API
Insert 2–3 photos in the product description
5. Publication via Prom API
The system should automatically create a product in the status of draft, including:
Title UA/RU
Short description UA/RU
Full HTML description UA/RU
Attributes (mapping by categories)
Search queries UA/RU
Price + old price (fake discount)
Category selection
Photos
Availability
Warehouse regions
Article
6. Product Condition Logic
Product condition (New, Like New, Refurbished, Used)
Packaging condition (Perfect, Damaged, No Box)
The system should automatically adapt the description depending on the condition.
7. Google Sheets Update
After successfully creating a product, the system should record:
Prom ID
Publication status
Creation date
Errors (if any)
8. Performance
Process a minimum of 1000 products in one cycle
Stability under high volumes
Automatic retries on Amazon or Prom API errors
Minimal human involvement
9. Code and Infrastructure Requirements
Python
Running on Mac or Linux VPS
Comments in the code
Deployment instructions
Separate configuration files (Prom, OpenAI, Google API keys)
10. Prohibitions
Do not use emojis
Do not use special characters, except for punctuation
Do not copy Amazon text without regeneration
Do not create unstructured descriptions
11. Expected Results
Fully functioning system
Complete package of source code
Test on 50 products
Deployment in my working environment
Scalability to 5,000+ positions