Runtime Engineer

Runtime Engineer – Heterogeneous AI Compute | C++ | Up to $300k + EquityI’m working with a well-funded AI infrastructure startup building a hardware-agnostic compiler and runtime platform designed to break the industry’s dependence on hardware-specific software stacks.Founded by an exceptional team of compiler, systems and semiconductor engineers, the company is approaching the commercial launch of its inference platform.ResponsibilitiesDesign, build and improve the multi-target runtime at the centre of the AI compiler stackTake compiler-generated code and execute it efficiently across GPUs, CPUs and heterogeneous acceleratorsDevelop asynchronous and concurrent execution systems for performance-critical workloadsExplore scheduling, parallelisation and workload-partitioning strategiesPrototype new runtime capabilities and evaluate them directly on target hardwareWork closely with compiler, kernel and product engineers to shape the wider runtime architectureHelp bring a greenfield AI systems platform into production at scaleIdeal experienceStrong modern C++ experience, ideally C++14 or newerExperience building compiler runtimes, GPU runtimes, systems software or other low-level execution infrastructureDeep understanding of asynchronous and concurrent programmingStrong knowledge of hardware architecture, including vector and scalar execution, registers and memory hierarchiesExperience with CUDA, ROCm or GPU compute libraries would be highly valuableGPU programming, HPC or distributed compute experienceFamiliarity with PyTorch, JAX, Triton or other modern ML frameworks