Description du poste
About the Role
We are seeking a DevOps Engineer with a strong software development background to join a distributed GPU compute platform project for a leading UK SaaS company in the fintech/enterprise planning space.
Project Context
You will be integrated into a senior engineering team focused on building a greenfield GPU-accelerated compute platform on AWS. While the central SRE team manages the underlying infrastructure, your primary responsibility will be to develop the necessary tooling on top of it.
Responsibilities
- GPU Benchmarking: Develop frameworks on AWS for scheduling benchmark runs, collecting and storing results, and enabling performance comparisons across different versions.
- Correctness Validation: Create tooling for automated testing of numerical accuracy of GPU compute outputs against reference results.
- Distributed Observability: Implement comprehensive observability across all platform services, including structured logging, distributed tracing (Pulsar), and performance metrics.
- HPC Contributions: Participate in broader High-Performance Computing (HPC) coding tasks alongside the engineering team.
Qualifications
- Development Skills: Proficient Python or Go developer with experience in writing production-level application code.
- Observability: Experience with observability tools such as Prometheus, Grafana, and distributed tracing.
- Cloud & CI/CD: Comfortable with AWS services (EC2, IAM, VPC) and CI/CD pipelines.
- HPC/GPU (Plus): Experience in HPC or GPU environments, including Slurm, compute clusters, and GPU workloads, is a significant advantage.
- Education: An engineering background from ENSIMAG, Centrale, INSA, X, or an equivalent institution is preferred.
- Language: Fluent in English, as the team is distributed across France and the UK.
Working Conditions
- Location: 100% remote, with an optional 1 day/week in London.
- Start Date: May 2026
- Contract Type: Freelance/Portage
- Compensation: Competitive, commensurate with experience.
