Open role
Research Engineer, Code Agents Infra
Mistral.ai
Palo Alto, California, United StatesPosted Aug 4, 2026 · Yesterday
Full-timeMid LevelHybridAI / ML
About this role
Mistral is seeking a Research Engineer to build and operate the infrastructure for their agentic models and coding assistants. You will design scalable systems for synthetic data generation, build fast training and RL execution environments, and maintain high-throughput execution engines. This role involves tackling large-scale sandboxing, data pipelines, training optimization, and low-latency orchestration across hybrid and multi-cloud environments.
What we are looking for
6- Design, deploy, and operate high-throughput sandboxing platform for LLM-generated code
- Architect and scale pipelines for synthetic code generation and agent data collection
- Optimize agent training codebases and distributed execution runtimes
- Implement low-latency orchestration using container warm pools and snapshot/restore technology
- Develop Kubernetes-native custom controllers for dynamic task routing
- Ensure strict multi-tenant isolation for untrusted agent code
