Senior Software Engineer - AI Compiler | 100% remote (Canadian-based) | $200,000 - $250,000 

The Company

My client is building the next-generation AI compiler for deep learning — a radically new infrastructure layer powered by LLM-based code generation and hardware-aware optimisation. As AI models get bigger, faster, and more specialised, traditional training workflows are becoming bottlenecks. They are solving that by rethinking how code is generated, optimised, and deployed for heterogeneous hardware environments. We're a fast-growing, innovation-obsessed team backed by leading VCs and AI researchers, working to redefine the compiler stack for AI at scale.

The Role

We are currently representing their search for a Senior Software Engineer – AI Compiler, to join their core compiler team. This is a remote role, but it must be is based in Canada (ideally Montreal or Toronto), though remote, but you must be based in Canada (ideally Montreal). Be at the heart of designing and building their AI compiler — an intelligent system that transforms model architectures into highly optimised training pipelines tailored for GPUs, AI accelerators, and other hardware targets. This is a rare opportunity to work at the intersection of compilers, ML frameworks, and LLM-driven agent systems. If you thrive in performance-critical environments and want to shape the infrastructure behind next-gen AI, this role is for you.

Responsibilities

• Collaborate with compiler engineers, ML researchers, and LLM systems teams to co-design Yasp’s intelligent AI compiler.
• Design and implement compiler passes for optimising model training workflows.
• Develop runtime optimisations and backend support for diverse hardware targets, including NVIDIA, AMD, and custom accelerators.
• Work with LLM-based agents to drive kernel code generation, schedule tuning, and operator selection.
• Analyse performance bottlenecks across the training stack (framework, kernels, memory layout) and apply low-level optimisations.
• Contribute to internal tooling that automates benchmarking, profiling, and performance validation.
• Maintain clean, production-level code with a strong emphasis on correctness, modularity, and testability.

Requirements

• Experience with AI compilers such as TVM, MLIR, XLA, or torch.compile.
• 5+ years of experience in systems programming, compilers, or ML infrastructure development.
• Strong programming skills in Python and C/C++.
• Hands-on experience with CUDA, LLVM, Triton, or similar codegen stacks.
• Familiarity with deep learning frameworks like PyTorch, TensorFlow, ONNX, or TensorRT.
• Experience with performance tuning and low-level optimisation on GPU, CPU, or AI accelerators.
• Solid understanding of machine learning training pipelines and model internals.
• Comfortable working in a Linux environment with Git, Docker, and CI/CD tools.

Preferred Qualifications

• Understanding of automatic kernel fusion, graph-level optimisation, or memory layout tuning.
• Familiarity with LLM agent systems and their role in program synthesis or model compilation.
• Contributions to open-source compiler, AI infrastructure, or ML optimisation projects.
• Exposure to JAX, Triton, or IR-based transformations for neural networks.
• Experience with profiling tools like Nsight, PyTorch Profiler, or Perfetto.

Benefits

• Competitive salary of $200K–$250K CAD, depending on experience.
• Generous stock option package.
• Comprehensive health and dental coverage.
• Flexible working hours and remote-friendly culture.
• Dedicated budget for conferences, research materials, and professional growth.
• Opportunity to work on frontier problems in AI infrastructure with some of the brightest minds in the field.

This is your chance to help rewrite the rules of AI performance. You won’t be building another abstraction layer, you’ll be working deep in the stack, crafting the compiler that tomorrow’s largest models will rely on. Join a mission-driven team with a bias for speed, rigor, and radical rethinking.

#welcometowhitebay

#michaelperry