Research Engineer, Code Agents Infra

Job summary

Responsibilities

This role focuses on building and operating the end-to-end execution, training, and data infrastructure that powers Mistral’s agentic models and coding assistants. You will be a core contributor to our agent research stack: designing scalable systems for synthetic data generation, building…

• Large-Scale Sandboxing Infrastructure: Design, deploy, and operate our high-throughput sandboxing platform, executing LLM-generated untrusted code across over 1 million isolated…
• Agent Data Generation Pipelines: Architect and scale high-throughput pipelines for synthetic code generation, agent trajectories, rollouts, and self-play data collection to power…
• Training Codebase & Systems Optimization: Optimize agent training codebases and distributed execution runtimes (PyTorch, Ray, SLURM/Kubernetes) to minimize multi-step rollout…
• Low-Latency Orchestration & Warm Pooling: Reduce sandbox cold-start times to sub-second levels using container warm pools, snapshot/restore technology (e.g., CRIU, microVMs), and…
• Multi-Cluster Queueing & Resource Allocation: Implement Kubernetes-native custom controllers, CRDs, and queuing systems to dynamically route short-lived evaluation, synthetic…

What they're looking for

• 4+ years of experience in Systems Engineering, Distributed Systems, Cloud Infrastructure, or MLOps supporting LLM/RL workloads.
• Data & Pipeline Engineering: Proven experience building high-throughput data processing and generation pipelines for large-scale datasets (e.g., Ray, Spark, custom distributed…
• Deep experience with Kubernetes & Container Tech: Strong expertise writing custom K8s operators/controllers, managing Linux cgroups/namespaces, and optimizing Docker image layers…
• High-Performance Software Engineering: Advanced proficiency in Python, Go, C++ or Rust, with a track record of profiling and optimizing high-performance ML or backend systems…
• Sandboxing & Isolation Technologies: Hands-on experience with lightweight virtualization, container runtimes, or WASM (e.g., Docker, gVisor, Firecracker).

Tech stack

• PyTorch
• Ray
• Kubernetes
• Spark
• Docker
• Python
• Go
• C++