We’re teaming up with one of the leading names in AI, known for pushing the boundaries of what’s possible with large-scale generative models and next-gen cloud infrastructure. This is a rare opportunity to step into a Staff Machine Learning Engineer role and play a key part in shaping the platforms powering millions of users across the globe. You'll be joining a team of exceptional researchers and engineers, all passionate about advancing the field and delivering world-class AI experiences. Location: London, Oxford Street (hybrid; onsite in London once a week) Rate: £1000 – £1200 per day, outside IR35 Start date: ASAP, 12-month contract What you'll be doing • Leading the development of scalable, reliable systems for training and fine-tuning transformer-based models • Optimising inference pipelines for real-time applications — aiming for low latency and high throughput • Exploring and applying advanced fine-tuning methods like LoRA, prefix-tuning, and adapters • Tuning performance across GPUs and systems using tools like DeepSpeed, Triton, TensorRT, and even custom kernels • Working closely with research, platform, and product teams to deliver new features and enhance the developer experience • Identifying and resolving bottlenecks through profiling, benchmarking, and performance tuning • Helping to define best practices for building, testing, and maintaining production ML services and APIs • Mentoring other engineers and helping to foster a culture of technical excellence and innovation What our client is looking for • 7+ years of experience building and deploying large-scale ML systems in production • Strong Python and PyTorch skills, with a deep understanding of transformers, LLMs, and multimodal models • Hands-on experience with distributed training frameworks like DeepSpeed or FSDP • Solid background in GPU programming (CUDA, ROCm) and inference optimisation • Practical experience with parameter-efficient fine-tuning techniques in real-world applications • Familiarity with container orchestration tools (Kubernetes, Kubeflow) and cloud-native environments • Knowledge of serving frameworks like Triton, vLLM, or similar • Clean, maintainable coding style and a strong testing discipline • Great communication skills and a collaborative mindset Your recruitment consultant Cameron Dalziel is a recruitment specialist in assembling teams in data, AI, design, and technology across Europe. He engages with top talent and is committed to providing a high quality service that delivers results. Connect with Cam to discuss this role.