Jobs

As a Model Behavior Architect on the Function Calling team at Mistral AI, you will define and measure how LLMs use tools and orchestrate complex agentic workflows. This role involves working with the Science team to establish standards for function calling, from accurate parameter selection to error recovery. You will design and implement evaluations, identify and fix edge case behaviors, and develop robust evaluation pipelines for model candidates. We are looking for experts in model evaluation, policy writing, and creating eval pipelines for tool use and function calling, who are passionate about tackling cutting-edge research challenges.