Handelsregister AppDocumentation

Extract structured company data with intelligent processing

Two-Tier Processing

Instant basic discovery via XML parsing (seconds) with optional deep research pipeline (minutes to hours) - users get immediate value with selective depth

Stateless Worker Architecture

Database-driven job orchestration with horizontal scaling, complete observability, and graceful error recovery for reliable document processing

Intelligent Data Extraction

Machine-readable XML for 100% accuracy, hash-based deduplication, and semantic chunking with RAG-based chat for comprehensive company research

Mono-Repo Development

TypeScript frontend and Python workers with shared database schema, automated type generation, and coordinated deployments via single Git workflow

Cost-Optimized Processing

SI XML eliminates OCR costs for basic data, selective deep research only when needed, and smart caching strategy with 7-day freshness

Enterprise-Ready Architecture

Multi-tenant organization support, role-based access control, comprehensive audit trails, and monitoring across four functional database schemas