Marsstein integrated the MATVIS AI Firewall into their platform architecture, placing it between their AI agents and the cloud AI models they rely on.
Every request from Marsstein’s AI agents to external models passes through MATVIS first. When an agent processes a client’s data processing records – containing names, contact details, data categories, and legal bases – MATVIS detects and pseudonymizes all personal data before it reaches the AI provider. The AI model works with sanitized data. When the response comes back, MATVIS re-identifies the pseudonyms so the output is complete and accurate.
Marsstein’s AI agents process data from multiple clients simultaneously. MATVIS ensures no client’s personal data ever leaves the platform in the clear. Because MATVIS uses deterministic machine learning models, the same input always produces the same classification – Marsstein can demonstrate exactly how data protection works to their own clients and auditors