ML Systems Engineer. Forlais Careers. Remote, UK hours.
Forlais. ML Systems Engineer. Build and run the training, evaluation, and dataset infrastructure that supports Forlais Research programmes and the EvaEsi.
Build and run the training, evaluation, and dataset infrastructure that supports Forlais Research programmes and the EvaEsi platform.
An engineering role on the systems that researchers and platform engineers use every day. The work spans training orchestration, evaluation harnesses, dataset tooling, and the operational reliability of the research and platform stack.
Responsibilities
Operate and extend the training orchestration stack across GPU clusters.
Maintain evaluation harnesses used by research staff and the EvaEsi release pipeline.
Own dataset tooling: ingestion, lineage, governance hooks aligned with FOR-POL-205.
Provide engineering review for research-engineering changes that touch shared infrastructure.
Participate in incident response on the research stack.
What we look for
Strong systems engineering background, ideally in Python and one of Rust / Go / C++.
Experience operating GPU clusters and distributed training jobs.
Comfort with observability, reproducibility, and change control.
Familiarity with at least one of: Kubernetes, Slurm, Ray, or comparable schedulers.