Mistral AI

Build the secure sandbox infrastructure that runs AI agents and reinforcement-learning workloads at massive scale. You will use Go and deep Linux systems expertise to deliver isolated, low-latency environments across Mistral’s sovereign AI cloud.

Build the storage foundation for Mistral’s sovereign AI cloud, including block, filesystem, and S3-compatible object storage. You’ll design production services in Go and own data-plane capabilities such as replication, tiering, backup, recovery, and caching at petabyte scale.

Build and own Mistral’s Kubernetes product for a sovereign European AI cloud, from bare-metal provisioning through GPU scheduling and control-plane reliability. You’ll write production Go, shape the platform architecture, and help scale infrastructure toward a gigawatt of compute.

Build the Go-based automation that turns newly installed datacenter hardware into accepted, customer-ready AI clusters. You will work across bare-metal provisioning, GPU fleets, networking, firmware, validation, and software deployment to reduce acceptance time from weeks to hours.

Build Mistral’s serverless containers platform for AI workloads, enabling developers to run CPU- and GPU-backed applications with fast cold starts and elastic scaling. You’ll own core scheduling, routing, isolation, and container infrastructure on Mistral’s sovereign cloud.

Build evaluation systems that help enterprise customers measure, improve, and trust large language model and agentic AI solutions. This role combines hands-on engineering, customer collaboration, and applied AI evaluation.

Build and operate a high-scale sandboxing platform that securely runs untrusted code across Kubernetes clusters and hyperscale infrastructure. This senior software engineering role focuses on distributed systems, virtualization, performance, and reliable platform design—not ML research.

Build and operate web-scale search infrastructure, data pipelines, knowledge services, and model-facing tools for Mistral’s research platform. This role supports large-scale data generation and AI training across hundreds of terabytes of storage and thousands of CPU cores.