26. 📷 Product Tagger

Overview

Computer vision service that automatically assigns category and attribute tags to product images for faster cataloging and search filtering.

Primary Use Cases

• Inventory teams mass-tagging new SKUs
• Marketplace platforms enriching listings
• Fashion retailers identifying clothing attributes

Key Features

• Multi-label classification: categories, colors, styles
• Bounding-box suggestions for attributes (e.g., collar, pocket)
• Batch-upload support
• Custom taxonomy builder

Tech Stack

• Frontend: React + TypeScript
• Backend: FastAPI (Python) + Go worker for batch jobs
• AI Models: resnet50 finetuned with torchvision, or Hugging Face vision-encoder-decoder pipelines
• Storage: S3 for images; PostgreSQL for tags

Architecture

1. Image Ingest: store uploads → queue tasks
2. Inference Worker: generates tags via CV model