Project Description
I need a complete n8n workflow that runs from the moment a new saree image lands in my Google Drive to the instant a promotional broadcast reaches customers on WhatsApp.
Source & trigger
• Watch a specific Google Drive folder for every new image drop. The design should stay flexible enough to swap in Amazon S3 later without ripping the flow apart.
Image generation
• Pass each source photo to ChatGPT image model 2 and return four views—front, side, walk, and close-up—while preserving fabric colours, motifs, drape and border details with pixel-level accuracy.
• If any frame fails validation (missing pattern, wrong trim), loop once for a clean retry, then log and flag.
Shopify upload
• Create or update the product, attach the four AI-generated assets, assign SKU, price, tags, and the right collection automatically.
• Capture the Shopify product URL and push it forward in the workflow.
WhatsApp broadcast
• Use Aisensi as the primary API to send a rich-media campaign that includes the product link and images. Keep the node structure abstracted so I can later flip to Twilio, WATI, or Gupshup with minimal refactor.
• Segment recipients by previous purchase behaviour pulled from Shopify metafields.
Reliability & observability
• Every critical node must carry retry logic, error-handled exits, and console + webhook logging.
• Write key metrics (items processed, failures, broadcast delivery stats) to a Google Sheet and a Slack channel.
Deliverables
- Exported n8n workflow (.json) ready to import
- Setup notes covering Google Drive auth, ChatGPT image model 2 credentials, Shopify keys, Aisensi token
- Quick-start video or screenshots proving a successful end-to-end run
- One revision round after initial hand-off to fine-tune edge cases
Acceptance criteria
1. Drop a sample image into Drive → four AI outputs appear on Shopify within 5 minutes.
2. WhatsApp broadcast reaches at least one test number with the correct product link and images.
3. Logs show zero unhandled errors after three consecutive test cycles.
The build must be modular, cleanly commented, and scalable enough to parallel-process future product lines beyond sarees. If you have already delivered a full n8n + API stack before, this should feel right up your alley.