Triton Compiler

Senior Compiler Engineer – Triton Backend

 Fully remote | USA preferred
 International considered
 Triton | LLVM | MLIR | AI Accelerators

Join a cutting-edge AI start-up building next-generation Triton compiler technology to unlock exceptional performance from custom AI hardware.

The company is developing high-performance compiler backends that allow Triton kernels to run efficiently across proprietary AI accelerators, helping hardware companies support modern machine-learning workloads without requiring engineers to rewrite kernels for every device.

They’re now looking for a Senior Compiler Engineer to help build this technology from the ground up.

What you’ll be doing:

• Designing and implementing custom Triton compiler backends
• Developing code-generation pipelines for proprietary AI accelerators
• Building hardware-aware optimisations for training and inference workloads
• Working on IR transformations, lowering and scheduling strategies
• Analysing performance bottlenecks and pushing kernels towards peak efficiency
• Collaborating with hardware architects and machine-learning engineers
• Helping shape the architectural direction of the compiler stack

What they’re looking for:

• Strong compiler development experience with LLVM, MLIR or similar technologies
• Experience with compiler backends, code generation and optimisation passes
• Strong low-level programming skills in C++, C or Rust
• A solid understanding of computer architecture, memory hierarchies and performance optimisation
• The ability to reason about hardware constraints and translate them into compiler strategies

Please message Harry Hansford for further information or a confidential discussion.