Staff AI Application Engineer – Heterogeneous Computing

This is a fully remote position, open to applicants in Belgium.

📋 Description

• Develop the model enablement platform and agent harness utilized by engineers and AI agents to deploy models.

• Create reusable tools for partitioning, compiling, executing, and profiling models.

• Provide real hardware environments and closed-loop executions that assess changes against reference benchmarks.

• Incorporate differential testing and regression detection into continuous integration.

• Manage the capture, analysis, and distribution of PyTorch or ONNX models across AIPU units, host CPU, and host accelerators.

• Construct placement tags that are user-readable and adjustable.

• Expand heterogeneous execution capabilities for large models, cascades, and multi-model pipelines.

• Enhance end-to-end performance through host/device overlap, zero-copy buffers, batching, multi-core scheduling, and multi-input models.

• Analyze device time, host time, and data movement, validating speed enhancements across the entire pipeline.

• Investigate critical customer models on Metis and Europa, transforming insights into platform enhancements.

• Collaborate with graph compiler, application, and runtime teams.

⛳️ Requirements

• Over 8 years of engineering experience in systems, compilers, runtimes, or performance, including ownership of system-level design.

• Practical experience with heterogeneous computing involving CPU, GPU, DSP, NPU, or FPGA.

• Proficient in reasoning about compute, memory bandwidth, latency, and data movement.

• Strong foundation in computer architecture, operating systems, and parallel programming.

• Knowledge of ML graph representations and compiler principles, including graph capture, lowering, operator fusion, partitioning, and memory planning.

• Proficient in C++ and Python.

• Daily use of AI coding agents in real engineering tasks, with the ability to assess and test their results.

• Preferred hands-on experience with MLIR or another compiler framework.

• Preferred experience with an inference runtime or execution engine.

• Preferred experience with GPU compute or SIMD optimization on Arm or x86 architectures.

• Preferred experience with techniques for executing large models.

• Preferred experience in building agent environments, evaluation harnesses, or verification frameworks.

🏝️ Benefits

• Pension plan.

• Extensive employee insurance options.

• Opportunity to acquire company shares.

• Flexible working arrangements.

• An open culture that fosters creativity and ongoing innovation.

• Collaborative ownership with freedom and responsibility.

• An equal opportunity and inclusive environment.