Site Reliability Engineer

Mistral•New York, NY

10h

About The Position

Mistral provides full-stack AI solutions: from frontier models to developer tools, applications, and compute. We partner with enterprises tackling the hardest problems—across high-stakes industries like finance, manufacturing, defense, healthcare, and the public sector—co-creating customized AI systems that they can run on their terms. We are a dynamic, collaborative team passionate about AI and its potential to transform society. Our diverse workforce thrives in competitive environments and is committed to driving innovation. Our teams are distributed between Europe, North America, Asia and the Middle East. We are creative, low-ego and team-spirited. The Role We are seeking highly experienced Site Reliability Engineers (SRE) to shape the reliability, scalability and performance of our platform and customer facing applications. You will work closely with our software engineers and research teams to ensure our systems meet and exceed our internal and external customers' expectations.

Requirements

Master’s degree in Computer Science, Engineering or a related field
7+ years of experience in a DevOps/SRE role
Strong experience with cloud computing and highly available distributed systems
Exposure to site reliability issues in critical environments (issue root cause analysis, in-production troubleshooting, on-call rotations...)
Experience working against reliability KPIs (observability, alerting, SLAs)
Hands-on experience with CI/CD, containerization and orchestration tools (Docker, Kubernetes...)
Knowledge of monitoring, logging, alerting and observability tools (Prometheus, Grafana, ELK Stack, Datadog...)
Familiarity with infrastructure-as-code tools like Terraform or CloudFormation
Proficiency in scripting languages (Python, Go, Bash...) and knowledge of software development best practices
Strong understanding of networking, security, and system administration concepts
Excellent problem-solving and communication skills
Self-motivated and able to work well in a fast-paced startup environment

Nice To Haves

experience in an AI/ML environment
experience of high-performance computing (HPC) systems and workload managers (Slurm)
worked with modern AI-oriented solutions (Fluidstack, Coreweave, Vast...)

Responsibilities

Design, build, and maintain scalable, highly available and fault-tolerant infrastructures to support our web services and ML workloads
Make sure our platform, inference and model training environments are always highly available and enable seamless replication of work environments across several HPC clusters
Operate systems and troubleshoot issues in production environments (interrupts, on-call responses, users admin, data extraction, infrastructure scaling, etc.)
Implement and improve monitoring, alerting, and incident response systems to ensure optimal system performance and minimize downtime
Implement and maintain workflows and tools (CI/CD, containerization, orchestration, monitoring, logging and alerting systems) for both our client-facing APIs and large training runs
Participate occasionally in on-call rotations to respond to incidents and perform root cause analysis to prevent future occurrences
Drive continuous improvement in infrastructure automation, deployment, and orchestration using tools like Kubernetes, Flux, Terraform
Collaborate with AI/ML researchers to develop and implement solutions that enable safe and reproducible model-training experiments
Build a cloud-agnostic platform offering an abstraction layer between science and infrastructure
Design and develop new workflows and tooling to improve to the reliability, availability and performance of our systems (automation scripts, refactoring, new API-based features, web apps, dashboards, etc.)
Collaborate with the security team to ensure infrastructure adheres to best security practices and compliance requirements
Document processes and procedures to ensure consistency and knowledge sharing across the team
Contribute to open-source projects, research publications, blog articles and conferences