Case Study: the Czech Ministry of Labour and Social Affairs achieves a unified AI-ready data dictionary with Adastra

Ministry of Labour and Social Affairs maps 30 years of data in days with Adastra

The Ministry of Labour and Social Affairs (MPSV), a Czech government body, faced the immense challenge of mapping and understanding its highly complex data environment, which consisted of dozens of databases, over 13,000 tables, and 100,000 columns accumulated over 30 years. Partnering with vendor Adastra, the ministry sought to create a unified data dictionary to improve digital services and enable future AI projects, a task that would have taken years to complete manually.

Adastra used generative AI and large language models (LLMs) to automatically classify tables and columns, detect duplicate data, and create a central business data dictionary. This solution processed a single schema in just 5-15 hours instead of weeks, removed 25% of duplicate data, and provided a validated, clear map of all information. The resulting dictionary significantly reduced employee onboarding time and provided a foundation for analytics, reporting, and future AI tools.

Ministry of Labour and Social Affairs (mpsv)