Research Engineer, Mistral AI
Code Agents Infrastructure
More information about location
Office located in Palo Alto, CA
Generative AI model developers
Open for applications
Generative AI model developers
501-1000 employees
Open for applications
More information about location
Office located in Palo Alto, CA
501-1000 employees
Company mission
To make frontier AI ubiquitous, and to provide tailor-made AI to all the builders.
Role
Who you are
- 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 queues)
- Deep experience with Kubernetes & Container Tech: Strong expertise writing custom K8s operators/controllers, managing Linux cgroups/namespaces, and optimizing Docker image layers and distribution systems
- 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 codebases
- Sandboxing & Isolation Technologies: Hands-on experience with lightweight virtualization, container runtimes, or WASM (e.g., Docker, gVisor, Firecracker)
- Queueing & Scheduling: Deep familiarity with task queue systems, resource schedulers, and low-latency queuing architectures for high-volume, short-lived workloads
- Comfort with Ambiguity: Passion for working directly alongside AI researchers to rapidly turn frontier agent ideas into scalable, production-grade infrastructure
What the job involves
- 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 ultra-fast training and RL execution environments, and maintaining high-throughput execution engines
- You will tackle the engineering challenges at every step of the agent lifecycle: from orchestrating 1M+ concurrent and short-lived sandboxes for untrusted code execution to optimizing agent training codebases, distributed trajectory collection pipelines, and dataset processing workflows across massive hybrid and multi-cloud clusters
- Large-Scale Sandboxing Infrastructure: Design, deploy, and operate our high-throughput sandboxing platform, executing LLM-generated untrusted code across over 1 million isolated environments concurrently for model evaluation and interactive RL environments
- Agent Data Generation Pipelines: 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
- Training Codebase & Systems Optimization: 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
- 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 optimized image delivery layers across hybrid clusters
- Multi-Cluster Queueing & Resource Allocation: Implement Kubernetes-native custom controllers, CRDs, and queuing systems to dynamically route short-lived evaluation, synthetic data, and agent execution tasks across diverse hardware fleets
- Isolation, Security & Security Boundary: Ensure strict multi-tenant network and process isolation for untrusted agent code using container/sandboxing runtimes (e.g., gVisor, Firecracker) and default-deny network postures
- Operational Excellence: Maintain high availability, telemetry, and automated self-healing across millions of transient jobs while participating in on-call rotations for critical agent training and execution pipelines
Our take
Founded by Facebook AI researchers, Mistral AI focuses on developing open-source generative AI models to "make AI useful." It emerged in early 2023, quickly achieving a $260 million valuation.
Mistral features up to 176 billion parameters, and its open-source nature has sparked concerns about potential misuse. However, Mistral remains a formidable competitor in the AI landscape, challenging models from Google and Meta.
In 2025, Mistral secured a massive €1.7 billion funding round, raising its valuation to €11.7 billion. This Series B funding, led by European chip firm ASML, underscores significant interest in alternatives to ChatGPT and a push away from Silicon Valley dependence by European AI competitors.
Freddie
Company Specialist at Welcome to the Jungle
Company
Funding (last 2 of 5 rounds)
Sep 2025
$1.9bn
SERIES C
Jun 2024
$525.1m
SERIES B
Total funding: $3.0bn
Company benefits
- Competitive bonus structure
- Equity
- Opportunities for professional growth and development
Company HQ
Villette, Paris, France
Leadership
Timothee Lacroix
(Co-Founder & CTO)Previously PhD and Software Engineer Intern at Facebook and Visiting Undergraduate at CAlab UCSD
Guillaume Lample
(Co-Founder & Chief Scientist)Research Student and Scientist at Facebook AI and Research Intern at Jane Street Capital
Arthur Mensch
(Co-Founder & CEO)Previously Staff Research Scientist at DeepMind, Post-doc researcher at École normale supérieure, Visiting Researcher at NYU Courant Institute of Mathematical Sciences and PhD Student at Inria
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 86 more jobs at Mistral AI