Digital Transformation of German Vehicle Registration
Bundeszulassung24 is a SaaS platform that fully digitizes the German vehicle registration process (new registration, re-registration, transfer, deregistration). As a licensed registration service, the platform handles all document and application processes for the customer — 24/7, fully online.
Monorepo Architecture
The project was structured as a Turborepo v2 + pnpm v9 monorepo with strict TypeScript 5.x. Four independent applications share a common database and type definitions:
| Application | Technology | Purpose |
|---|---|---|
| Web (Customer) | Next.js 14 PWA | Customer portal with service router and multi-step forms |
| API | Fastify 4 | REST API with JWT auth, Zod validation, 10+ route modules |
| Admin | Refine.dev + Ant Design | Full management with dashboard and SEO management |
| Operator | Refine.dev + Ant Design | Processing queue and customer communication |
22 Database Models
PostgreSQL 16 with Prisma 5 ORM — 22 tables and 13 enums forming a comprehensive data model:
- User (25 fields) — 5-tier role system (super_admin → customer)
- Submission — Core business unit with 11-stage status flow
- Vehicle (23 fields) — VIN, plate, insurance, security codes
- Document — OCR metadata via Google Document AI
- Payment — Stripe integration with refund tracking
- SEO Models — SeoSettings, SchemaMarkup, AiVisibility, BrandLock, GeoPages
AI-Powered Document Recognition
Google Document AI integration for automatic extraction from vehicle documents (registration certificate, ID card, eVB). Confidence score system with automatic field population:
- ≥80% confidence → Automatic acceptance
- 50-79% → Customer review recommended
- <50% → Manual input
Infrastructure & Security
Single-server deployment with 8 Docker containers orchestrated by Docker Compose. Traefik v2.11 as reverse proxy with automatic Let's Encrypt SSL. 10+ security layers: JWT (8h access + 7d refresh), bcrypt 12 rounds, Redis rate limiting (100 req/min), Zod validation, GDPR/Cookie Consent Mode v2.
CAFE SEO Strategy
7-layer SEO/AI visibility strategy using the CAFE Framework: GEO-optimized Bundesland pages, E-E-A-T signals, 8+ pages with Schema.org markup, llms.txt for AI visibility, BrandLock brand consistency, and 35+ canonical URLs with hreflang.