DevOps Engineer, HPC

About the Role

Join Mistral AI's fast-growing Compute team as a DevOps Engineer, HPC, to build, scale, and automate one of Europe's largest AI infrastructures. You will be instrumental in engineering fault-tolerant and reliable platforms that support both internal operations and customer-facing AI services.

Responsibilities

• Design, build, and operate a scalable Kubernetes-based platform to host large-scale AI and HPC workloads, ensuring high performance, reliability, and security.
• Own the full lifecycle of cluster management, from bootstrapping and provisioning to global operations, by integrating and developing necessary software components.
• Drive infrastructure innovation by designing workflows, tooling (scripts, APIs, dashboards), and CI/CD pipelines to optimize system reliability, availability, and observability.
• Champion a zero-trust security model, strengthening IAM, networking (VPC), and access controls to safeguard the platform.
• Develop user-centric features that simplify operations for both sysadmins and end customers, reducing friction in daily workflows.
• Lead incident resolution with rigorous root-cause analysis to prevent recurrence and improve system resilience.

Requirements

• Successful experience in an Infrastructure Engineering role (Software Engineer, DevOps, SRE, Platform).
• Strong proficiency in software development (preferably Golang) and knowledge of software development best practices.
• Deep understanding of Kubernetes internals and hands-on experience with containerization and orchestration tools (Docker, Kubernetes).
• Familiarity with infrastructure-as-code tools like Terraform or CloudFormation.
• Knowledge of monitoring, logging, alerting, and observability tools (e.g., Prometheus, Grafana, ELK, Datadog).
• Exposure to highly available distributed systems and site reliability issues in critical environments.
• Experience working against reliability KPIs (observability, alerting, SLAs).
• Excellent problem-solving and communication skills.
• Self-motivation and ability to thrive in a fast-paced startup environment.

Nice-to-Haves

• Experience with HPC workload managers (Slurm) and distributed storage systems (Lustre, Ceph).
• Demonstrated history of contributing to open-source projects (e.g., code, documentation, bug fixes, feature development, or community support).