Research Engineer, Code Agents Infra

Join Mistral, a leading AI solutions provider, as a Research Engineer focused on building and operating the infrastructure that powers agentic models and coding assistants. You will tackle engineering challenges at every step of the agent lifecycle, from sandboxing to training codebase optimization.

• Design, deploy, and operate a high-throughput sandboxing platform for executing LLM-generated untrusted code across over 1 million isolated environments concurrently.
• Architect and scale high-throughput pipelines for synthetic code generation, agent trajectories, rollouts, and self-play data collection to power post-training and RL loops.
• Optimize agent training codebases and distributed execution runtimes (PyTorch, Ray, SLURM/Kubernetes) to minimize multi-step rollout overhead, improve GPU utilization, and eliminate scaling bottlenecks.