ML Compiler Engineer, Senior

As a Senior ML Compiler Engineer at d-Matrix, you will primarily focus on developing the compiler backend. This involves solving pivotal algorithmic compiler challenges, such as mapping low-level instructions to hardware resources. Additionally, you will work on model partitioning, resource allocation, memory management, and optimization tasks that aim to improve system performance, which are crucial for effectively leveraging AI capabilities.

The ideal candidate for the Senior ML Compiler Engineer role at d-Matrix should hold at least a Bachelor’s degree in Computer Science, with 4+ years of relevant industry experience. An MSCS degree is preferred, along with experience in modern compiler infrastructures like LLVM and MLIR. A capability to write production-quality code in modern C++ and familiarity with machine learning frameworks like TensorFlow and PyTorch is also essential.

In your role as a Senior ML Compiler Engineer at d-Matrix, you'll engage in projects focusing on the optimization and resource allocation in the compiler backend. You'll work on intricate tasks such as model partitioning, tiling, memory management, and enhancing system performance parameters like bandwidth and throughput in spatial architectures.

At d-Matrix, you can expect a collaborative and inclusive work environment focused on respect and direct communication. Our team values diverse perspectives and foster a culture where knowledge sharing is encouraged, allowing you to thrive within a supportive community. With a hybrid work model, you’ll enjoy flexibility while still engaging meaningfully with the team onsite in Toronto.

d-Matrix is committed to your professional growth as a Senior ML Compiler Engineer by providing opportunities for skill enhancement through collaboration with seasoned developers and engaging in challenging projects. Our culture emphasizes learning and development, allowing you to embrace challenges and expand your expertise in cutting-edge AI technologies.

When answering this question, highlight specific projects where you've utilized LLVM, detailing your role, the challenges faced, and the outcomes. It's essential to demonstrate both your technical skills and an understanding of how LLVM plays a crucial role in compiler optimization.

Provide a clear outline of the methods and algorithms you've employed for resource allocation in previous projects. Use examples to illustrate your thought process and demonstrate familiarity with concepts like memory management, latency considerations, and throughput optimization.

When discussing a challenging problem, share the context, the steps taken to resolve it, and the tools or methodologies utilized. Highlight any innovations or efficiencies you introduced during the process, showcasing your analytical skills and problem-solving capabilities.

Emphasize your commitment to writing production-level code by mentioning practices like code reviews, unit testing, and the use of CI/CD pipelines. Discuss any coding standards you adhere to, and explain how these practices contribute to the overall quality and reliability of the compiler.

Clarify the concept of model partitioning and its importance in compiler design. Discuss specific strategies for achieving effective partitioning, such as pipelined execution or data parallelism, and provide examples of how you’ve implemented these practices in your work.

Discuss your hands-on experience with these frameworks, highlighting any projects where you've incorporated them into your work. Elaborate on how you’ve used their features to improve compiler functionality or performance, demonstrating both your technical capability and understanding of real-world application.

Detail your approach to task prioritization, including methodologies like Agile or Kanban if applicable. Share examples of how you manage competing deadlines and requirements, emphasizing your organizational skills and ability to adapt to changing priorities.

Offer insights into specific optimization techniques you've employed, such as loop unrolling or inlining, and how they improved program performance. Use quantitative results when possible to showcase the effectiveness of your optimizations during your previous experiences.

Frame your response around a specific instance where you had to rapidly acquire knowledge on a new technology, detailing the steps you took to familiarize yourself, the challenges faced, and how you applied that knowledge effectively in your project.

Share strategies you employ to stay updated on industry trends, such as participating in relevant forums, attending conferences, reading technical journals, or engaging in open-source projects. Highlight specific resources that have significantly contributed to your professional development and knowledge in the compiler domain.