AI Compiler Engineer - Compilation - TVM - MLIR - Deep Learning - Neural network Compiler - Remote

AI Compiler Engineer - Compilation - TVM - MLIR - Deep Learning - Neural network Compiler - Remote

Our client is seeking a Staff AI Compiler Engineer to lead the design and evolution of their cutting-edge neural network compiler. This is a key technical leadership role focused on architecting scalable, high-performance software solutions that power next-generation AI hardware.

Key Responsibilities

• Lead the architecture, design, and hands-on development of the neural network compiler and SDK.

• Develop advanced compiler optimizations tailored to in-memory computing architectures, with a focus on hardware-aware transformations.

• Drive complex, long-term initiatives to expand compiler support to emerging workloads (eg, generative AI) and future versions of the client's heterogeneous compute platform.

• Collaborate cross-functionally to align technical strategies, ensuring scalability, flexibility, and performance across the software stack.

• Oversee the entire compilation pipeline, proactively recommending improvements to enhance functional coverage, performance, and accuracy.

• Champion best practices in compiler development, particularly in testing, validation, and continuous integration.

• Provide technical mentorship to engineers at all levels, modelling excellence and fostering a collaborative, high-impact engineering culture.

Qualifications

• Master's or PhD in Computer Science, Electrical Engineering, or a related field.

• 5+ years of software engineering experience, with at least 2 years focused on AI or deep learning systems.

• Deep expertise in AI compiler development, particularly in areas such as:

  • Workload partitioning for heterogeneous platforms

  • Scheduling and tiling optimizations

  • Multi-level memory hierarchy management

  • Synchronization and runtime integration

• Solid understanding of computer architecture and heterogeneous compute systems.

• Hands-on experience with open-source compiler frameworks (eg, TVM, MLIR, Glow).

• Familiarity with neural network quantization techniques is a strong plus.

• Proficiency with PyTorch and a solid grasp of modern deep learning frameworks.

• Strong coding skills in both C++ and Python