Compiler Engineer

About the Company

Charterhouse is partnering with a rapidly growing technology innovator that specializes in designing and developing cutting‑edge, customized server hardware optimized for artificial intelligence and machine learning applications. Their mission is to accelerate AI initiatives globally by delivering high‑performance, scalable, and energy‑efficient infrastructure.

We are seeking an experienced Compiler Engineer to develop MLIR dialects and lowering pipelines for advanced AI accelerator/NPU platforms. The role involves transforming machine‑learning graphs from frameworks such as ONNX and PyTorch into optimized kernels, focusing on fusion, tiling, and vectorization, and profiling to meet stringent latency, throughput, and power targets.

• Design, implement, and maintain MLIR dialects and lowering pipelines for AI accelerator hardware.
• Transform computational graphs from ONNX, PyTorch, and Tensor Flow into optimized kernel code, applying fusion, tiling, and vectorization techniques.
• Profile, benchmark, and optimize compiler output to achieve target performance and power metrics.
• Collaborate closely with hardware architects, runtime engineers, and framework developers to co‑design compiler features and ensure integration.
• Build a portable compiler stack that supports multiple hardware generations and maintain backward compatibility.
• Document design decisions, usage patterns, and performance trade‑offs for internal and external stakeholders.

• Minimum five years of experience in compiler engineering or high‑performance systems.
• Proficiency in modern C++ and comfortable using Python for tooling and scripting.
• Hands‑on experience with MLIR or LLVM, including dialect design, lowering, and optimization.
• Strong understanding of compiler internals such as intermediate representations, scheduling, vectorization, and loop transformations.
• Familiarity with major AI/ML frameworks (PyTorch, ONNX, Tensor Flow) and core operations like tensor manipulation, matrix multiplication, quantization, and dynamic shapes.
• Bachelor’s degree in Computer Science, Computer Engineering, or related field (advanced degree preferred).

• Experience targeting NPUs or AI accelerator hardware.
• Contributions to open‑source compiler projects such as MLIR, LLVM, Torch‑MLIR, or ONNX‑MLIR.
• Knowledge of runtime systems, scheduling, resource sharing, and memory movement engines.

• Seniority Level: Mid‑Senior
• Employment Type:
  
  Full‑Time
• Job Function:
  Information Technology
• Industry: Information Services
• Location:
  
  Dubai, United Arab Emirates

Competitive compensation, opportunities for professional growth, and the chance to shape next‑generation AI hardware and software ecosystems.