AI Evaluation, Guardrailing & EU AI Act Advisory

I help teams build and ship LLM systems they can trust, especially in regulated and public-sector settings where correctness, traceability, and compliance matter.

Typical engagements:

• Designing and running LLM evaluation and red-teaming programmes

• Specifying and implementing guardrails for production LLM/RAG systems

• RAG quality assurance: groundedness, consistency, contradiction detection

• Prompt management policy and PromptOps practices

• EU AI Act readiness reviews and LLM governance

I additionally bring a multilingual perspective (EN/DE/ES/FR/NL) useful for cross-lingual, EU-wide deployments.