Nota AI
AI Inference Optimization Engineer
Seoul, South KoreaPosted 24 days ago
What you'd do
- Analyze mismatches between AI models and target hardware (NPU,GPU,CPU) and track root causes to resolution
- Design and implement graph-level optimization passes (op fusion,folding,decomposition,replacement)
- Explore and experiment with optimization opportunities considering both model structure and hardware characteristics
What they want
- 2+ years relevant post-graduation practical experience,or Masters degree or higher
- PyTorch,ONNX,Python,Linux,Git practical experience
- End-to-end experience following AI model training through actual device execution
Nice to have
- Graph IR-level model analysis and transformation (TensorRT,TFLite,ExecuTorch)
- NPU or custom accelerator target model optimization experience
- Deep understanding of GPU/NPU kernels