Modern Auto
A serverless automotive marketplace built around fast inventory discovery, secure administration, and an optimized image-heavy browsing experience.
- My role
- Full-stack developer · Marketplace architecture, inventory search, role-based access, data modeling, and performance
- Technology
- Next.js
- PostgreSQL
- Prisma ORM
- NextAuth.js
- Supabase
- Tailwind CSS
- Sharp
01
The problem
Automotive marketplaces need to make large, image-heavy inventories easy to search while protecting every administrative workflow. The product needed flexible discovery across vehicle attributes, reliable favorites, and a media pipeline that would stay fast for mobile visitors.
02
Architecture and decisions
I used a serverless Next.js architecture with PostgreSQL and Prisma for structured inventory data. Search inputs map to composable database filters across more than six vehicle attributes, while NextAuth and role checks secure administrative and user operations across the API surface. Next.js Image, Sharp, lazy loading, and WebP conversion reduce the cost of high-resolution inventory photography.
03
Key features
- Inventory discovery across 6+ vehicle filters
- Saved favorites and streamlined vehicle-detail journeys
- Role-based access control across 15+ API endpoints
- Protected administration and user permissions with NextAuth.js
- Responsive image sizing, lazy loading, and WebP optimization
04
Outcomes
- 01
Supported an automotive marketplace serving more than 1,000 monthly visitors.
- 02
Reduced the average time needed to reach a relevant listing by approximately 3×.
- 03
Improved page-load performance by approximately 45% and reduced image bandwidth by approximately 65% in Lighthouse testing.
Metrics are included only where they were measured or documented.
Inside the product
Interface highlights
