Model Behavior Architect – Safety & Evaluation

Mistral is seeking a Model Behavior Architect for the Safety team to shape how large language models reason, translate, and use tools. You will work with the Science team to translate research into actionable improvements.

You will design evaluations, data guidelines, and synthetic tests, address edge cases, and build robust evaluation pipelines in collaboration with AI Scientists to ensure high performance and real-world utility.

#J-18808-Ljbffr…