Senior Machine Learning Engineer - Research Optimisation

Senior ML Engineer focused on research enablement and performance, building production-ready pipelines that bridge research and production. Responsibilities include hardening experimental models (containerisation, tests, CI/CD), profiling PyTorch training jobs and improving GPU utilisation, developing reusable libraries and services in a monorepo, building inference services and SDKs, creating multi-variant rollout frameworks, establishing CI/CD workflows and reproducible builds, and enhancing observability and reliability across training and inference workloads. Collaborate with researchers and engineers to accelerate adoption, optimize end-to-end ML workflows, and reduce time and cost per experiment. Nice-to-have skills include CUDA kernel work, TorchScript/Triton, distributed training frameworks (FSDP, DDP, DeepSpeed), high-performance storage (Weka, Vast, Lustre), feature-flag experimentation platforms, MLOps practices, and experience with multimodal/LLM-adjacent tooling. The role involves Kubernetes orchestration, storage and data pipelines, and close collaboration with researchers, DevOps, and platform engineers to ship ML innovations at scale.