PassPad
Full-stack Re-Commerce platform for video games
AI-powered product recognition • Headless Shopify • Production-ready
About the project
PassPad is a re-commerce startup I founded with my cousin to digitize the used video game market. Our platform solves a classic market problem: the high manual effort required to value collections.
As a full-stack developer and analyst, my goal was to consistently replace manual processes with technology. We built a highly automated pipeline that dramatically reduces admin work using computer vision and smart pricing algorithms. Today the system manages over 10,000 items largely autonomously—tasks that used to take hours now run in the background.
The vision
PassPad unites two worlds in one modern platform:
Buyback
Users sell used games and consoles—with an AI scanner for fast recognition and a transparent checkout flow.
Shop
Used products are sold through a headless Shopify shop—SSR, SEO-optimized, and scalable.
Technology stack
Next.js 16
App Router
React 19
Server Components
TypeScript
End-to-end types
PostgreSQL
Prisma ORM
Shopify
Storefront API
Gemini
AI integration
PostHog
Analytics
Google Tag Manager + Analytics
Tracking & analytics
Feature spotlight
AI scanner
The signature feature: less typing, faster offers, better conversion.
- ✓ Multi-image upload with progress states
- ✓ Batch-based async processing
- ✓ Admin mode & A/B testing controls
- ✓ Server-side asset handling (S3)
Headless shop
Shopify Storefront API with typed GraphQL queries, SSR, and SEO foundations.
In our shop we sell processed video games and consoles.
Admin pipeline
Dedicated admin tooling for managing purchases, products, coupons, and more.
In our admin area we can manage all purchases, products, coupons and more.
Architecture
Clean architecture: business logic and data access clearly separated from the UI.
Actions
Server Actions with auth, validation, revalidation
Use cases
Business logic orchestrated (e.g. upload workflows)
DB layer
Encapsulated data access layer
Performance & search
🔍 PostgreSQL full-text search
- • GIN indexes for multilingual search
- • Composite indexes for filter combinations
- • Optimized queries with EXPLAIN ANALYZE
📊 Performance tracking
- • PostHog events (duration_ms, results_count)
- • A/B testing of components (e.g. AI scanner)
- • Measurable instead of gut feeling