AI Runtime Security, Guardrails & Injection Defense

European Commission-backed open-source testing platform for LLMs.

Co-creator of COMPL-AI benchmark evaluating frontier models against EU AI Act.

Open-source LLM security CLI with automated red-teaming test suites.

Swiss AI security provider maintaining the Gandalf threat intelligence dataset.

Enterprise observability platform providing explainability and live guardrails.

Self-hosted AI control plane enforcing real-time compliance policies on LLM inferences.