About Us
Keeping your AI applications on track
Miriam brings deep technical expertise in large language models, combined with years of experience building production AI systems for clients across legal publishing, news media, finance, and government.
At deepset, she led customer-facing engineering teams and guided organizations through LLM implementations - from product design and strategy to technical delivery of semantic search, RAG applications and agents.
Her work gave her firsthand insight into the growing compliance and control challenges companies face with AI systems. She understands both the technical complexity of evaluating LLM applications and the organizational realities of deploying them responsibly.
Mathis is a deeply technical product leader with over seven years spent building and scaling AI products for enterprises. At deepset, he spearheaded product development as VP of Product, growing their platform from initial customer deployments to a comprehensive AI development solution.
He designs and evaluates AI systems through close partnership with domain experts, learning how practitioners work and what standards matter in their environments.
Mathis has published research on large language models, won a Kaggle competition on AI-based text complexity assessment, and speaks at international AI conferences.
Our team is based in Berlin and we are excited to contribute to the European AI ecosystem.
Our Mission
AI is changing the way we operate. It automates routine tasks, allowing humans to reclaim headspace for activities that require their unique depth and creativity.
But careless AI deployment creates problems. When systems hallucinate or produce unwanted outputs, work quality degrades. Organizations are already dealing with the fallout.
We founded Blue Guardrails to help teams build AI responsibly. We help them tackle core technical challenges: preventing hallucinations, ensuring robustness, maintaining human oversight, and addressing fairness. Our approach is deeply embedded in the domain context of our customers.
The AI being built today will determine whether technology enhances or diminishes human capability. We exist to ensure it enhances—creating environments where AI makes work better, not just faster.