We are seeking a Senior/Staff AI Graph Compiler Engineer to design and optimize MLIR-based compiler components for efficient execution of AI workloads on cutting-edge heterogeneous hardware. You will collaborate with AI researchers, compiler engineers, and hardware architects across Europe to advance compiler technology within a fast-moving deep-tech startup environment at Axelera AI.
What You'll Do
- Design, implement, and maintain frontend and graph-level compiler components using MLIR
- Develop and optimize graph-level transformations such as operator fusion, constant folding, operator sinking, graph partitioning, and other performance-critical optimizations
- Extend and maintain MLIR dialects, passes, and infrastructure to support AI workloads
- Integrate and lower AI models from frameworks such as PyTorch, ONNX, and TensorFlow into internal compiler representations
- Collaborate with hardware and backend compiler teams to ensure efficient mapping of AI workloads to heterogeneous architectures
- Support and mentor team members in adopting and effectively using MLIR infrastructure
- Analyze model graphs and implement optimizations to improve performance, memory efficiency, and execution efficiency
- Contribute to the design and evolution of the overall compiler architecture and tooling
- Debug, profile, and improve compiler performance and correctness
What We're Looking For
- Master’s or PhD in Computer Science or a related technical field
- 3-5 years of experience in a Software Engineering role, thereof ideally 2 years of experience with Deep Learning frameworks and AI systems
- Experience with MLIR, including dialects, passes, and compiler infrastructure
- Solid understanding of compiler design principles and intermediate representations, especially graph-based IRs
- Experience working with AI model frameworks such as PyTorch (preferred), ONNX, or TensorFlow
- Proven experience implementing graph-level optimizations such as operator fusion, constant folding, graph partitioning, or similar transformations
- Strong programming skills in C++ and Python
- Experience working in collaborative engineering teams
- Excellent communication skills and willingness to share knowledge and mentor others
Nice to Have
- Experience working with custom AI accelerators or specialized hardware
- Background in computer architecture, especially heterogeneous systems (e.g., CPU + NPU, GPU, or dedicated accelerators)
- Experience with AI compiler stacks such as Torch-MLIR, TVM, XLA, Glow, or similar
- Experience optimizing AI workloads for performance and efficiency
- Familiarity with frontend model ingestion, graph lowering, and compiler pipelines
Technical Stack
MLIR, C++, Python, PyTorch, ONNX, TensorFlow, Torch-MLIR, TVM, XLA, Glow
Team & Environment
Join the compiler team at Axelera AI, part of a 220+ employee company with 49+ PhDs. Work in a collaborative team of engineers across Europe, working closely with AI researchers, compiler engineers, and hardware architects.
Benefits & Compensation
- Flexible working arrangement with office or remote options in Europe
- Attractive compensation package
- Pension plan
- Extensive employee insurances
- Option to get company shares
- Open culture supporting creativity and continual innovation
- Collaborative ownership and freedom with responsibility
- Equal opportunity and diversity commitment
- Warm and inclusive environment
Work Mode
Local-country roles available in Leuven, Belgium; Amsterdam, Netherlands; Eindhoven, Netherlands; Zurich, Switzerland; Florence, Italy; Milan, Italy; Bristol, United Kingdom; or remote within any European country (including UK). Relocation options to Italy (Florence or Milan) or Netherlands (Amsterdam or Eindhoven). Flexible working arrangement; priority given to candidates based in or interested in relocating to Belgium or Italy.
Axelera AI wholeheartedly embraces equal opportunity and holds diversity in the highest regard, committed to cultivating a warm and inclusive environment that empowers and celebrates every member of the team.
