In short
Infrastructure engineer at Mistral AI building the data backbone for frontier-model training and fine-tuning: owning scalable compute fleets, distributed storage and secure data access across on-prem and cloud-native Kubernetes, plus scheduler migrations, production pipelines and on-call for critical training jobs.
About the role
Mistral AI is building the backbone of its frontier-model training and fine-tuning ecosystem. This Data Infrastructure role focuses on scalable compute fleets, distributed storage, and secure data access for MLOps and research, across on-prem and cloud-native Kubernetes environments.
You will own migrations from legacy schedulers, implement production pipelines, and participate in on-call rotations for critical training jobs, with a hybrid/remote work option in EU/UK.
Requirements
- Experience with on-prem and cloud-native Kubernetes environments
- Experience with scalable compute fleets and distributed storage
- Secure data access for MLOps and research
- Own migrations from legacy schedulers
- Implement production pipelines
- Participate in on-call rotations for critical training jobs