Virtual Try-On — AI Fashion Platform

Full-Stack (FastAPI + Next.js)

About

Full-stack AI-powered virtual try-on platform for e-commerce — enabling customers to visualize clothing using image processing, with async task queues and real-time WebSocket updates.

My Role

Built the complete platform: FastAPI backend with SQLAlchemy ORM, Celery task queue for background image processing, Alembic migrations, and Next.js 14 frontend.

Core Capabilities

• AI-powered virtual clothing try-on with image processing
• Async task queue with Celery and Redis for background jobs
• Real-time WebSocket updates for processing status
• Database migrations with Alembic version control
• Product scraping module for catalog building
• Multipart form handling for image uploads

Architecture

• Python FastAPI with async HTTPX and WebSocket support
• SQLAlchemy ORM with PostgreSQL via psycopg2
• Celery with Redis backend for task processing
• Alembic for database migration management
• Next.js 14 with TypeScript frontend

Tech Stack