Senior AI Model Enablement Engineer - Heterogeneous Compute

Axelera AI; Belgique; Remote; EUR 120,000 - 150,000; Full time

Job summary

Axelera AI is seeking a Senior/Staff Software Engineer to build the model enablement platform and enable efficient execution across Metis/Europa AIPU, host CPUs, and accelerators. You will work on the frontend, graph compiler, runtime, and application pipeline, collaborating with AI agents in daily tasks.

The role blends architecture with hands-on development and welcomes experts returning to hands-on work. Remote-friendly with offices in Leuven and Europe-wide locations.

Qualifications

• 8+ years experience in systems, compilers or runtimes with design ownership.
• Experience with heterogeneous CPU/GPU/DSP/NPU/FPGA.
• Strong fundamentals in architecture, OS and parallel programming.
• Familiarity with ML graph representations and compiler concepts (graph capture, lowering, fusion).
• Proficient in C++ and Python.

Responsibilities

• Build the model enablement platform and agent harness for bringing up models across AIPU, CPU and accelerators.
• Own model ingestion: capture and split PyTorch/ONNX models across device units with placement tags.
• Extend partitioning to large-model pipelines and multi-model cascades with caching/offloading.
• Collaborate with teams to optimize end-to-end performance (overlaps, zero-copy, batching, scheduling).
• Study critical models and translate lessons into platform improvements.

Skills

8+ years engineering experience

Heterogeneous computing

Computer architecture

ML graph representations

C++ and Python

AI coding agents usage

MLIR / compiler frameworks

Inference runtimes

GPU compute / CUDA/OpenCL

Large-model execution techniques

Agent environments / verification

Tools

Torch / ONNX tools

Job description

Axelera AI is seeking a Senior/Staff Software Engineer to build the model enablement platform and enable efficient execution across Metis/Europa AIPU, host CPUs, and accelerators. You will work on the frontend, graph compiler, runtime, and application pipeline, collaborating with AI agents in daily tasks.

The role blends architecture with hands-on development and welcomes experts returning to hands-on work. Remote-friendly with offices in Leuven and Europe-wide locations.