Research Engineer, Inference Foundation

Join Mistral, a leading AI solutions provider, as a Research Engineer in the Inference Foundation team. In this hybrid role, you will optimize the inference stack at scale, make capacity elastic, and power the training of frontier models. You will work on inference engine and orchestration, performance and capacity at scale, and serving for frontier training. Ideal candidates have experience building and running ML/LLM services at scale, hands-on experience with inference engines, and familiarity with distributed serving architectures.

Responsibilities

• Optimizing the inference stack at scale, including feature development and fixes in the engine and orchestrator.
• Owning the release process for the serving stack, ensuring validated, regression-free releases through automated performance gates.
• Building the serving infrastructure that powers reinforcement learning and post-training for frontier models.