🚀 What's New in v2.0
Virtual Wardrobe has been rebuilt from the ground up to deliver faster performance, AI-powered features, and a modern, cloud-native experience.
Full Frontend Rebuild with Modern Stack
- Migrated to Next.js with TypeScript.
- Used Tailwind CSS for utility-first styling.
Backend Revamp with FastAPI
- Switched from Django to FastAPI for async APIs.
- Built as modular microservices.
- Used Pydantic for schema validation.
Authentication & Authorization
- Google OAuth login support.
- JWT-based secure authentication.
- Role-based access via Prisma enums.
AI-Powered Virtual Try-On
- CatVTON integration via fal.ai.
- Upload image and outfit to preview virtually.
Wardrobe Organization & Filtering
- Advanced filters for type, color, season, and category.
- Real-time wardrobe updates and sorting.
- Optimized uploads to GCP Cloud Storage.
Performance & Observability
- Redis caching for faster API response.
- Grafana Loki + Promtail for real-time logs.
- Structured logging for better traceability.
Cloud-Native Deployment
- Backend hosted on GCP Compute Engine.
- Neon serverless postgres with autoscaling.
- Static/media file handling on GCP Storage.
DevOps & Scalability Enhancements
- API response time improved by ~40%.
- Horizontal scaling enabled.
- Feature-first modular codebase.