For companies:
For professionals:
For everyone:
For developers:
Responsibilities
What you'll doDesign, architect, and ship large-scale foundation models with modern architectures (MoE, state-space models)Write custom CUDA kernels where off-the-shelf libraries fall shortScale distributed training and inference pipelines on cloud infrastructureOwn ML systems against strict SLOs/SLAs, and build internal tooling to accelerate the teamWhat we're looking forShipped a large-scale foundation model 0→1 at a high-growth AI startup or top-tier research labDeep, hands-on ML modelling and training expertise — this is weighted above raw systems skillStrong proficiency in both Python and CUDA C/C++, including custom kernel developmentDirect experience scaling distributed training/inference on AWS, Azure, or GCPDeep knowledge of a major DL framework (PyTorch ideal)A track record of delivering fast in ambiguous, early-stage environmentsLocationBased in London (office-first), or fully remote from BrazilComfortable working with a London-based team across time zonesThe detailsFull-timeTwo-stage process: a values conversation with the CEO and one deep technical session with the CPO — no take-home or live coding tasksA VC-backed AI startup in West London is building a novel foundation model for a genuinely first-of-its-kind product: fully automated, unsupervised software delivery in embedded control systems. They're scaling their technical team and hiring a Senior ML Engineer to own the foundation model end-to-end research through production.This is a hands-on, high-ownership role reporting close to the founders, in a direct, no-jargon, high-transparency environment. The team is London-based, and they're open to strong remote candidates based in Brazil.Stack: CUDA · PyTorch · Python · C++ · distributed training · GPU · ML infrastructure
This job post is closed.