I'm a computer vision engineer focused on building and deploying AI models on real-world devices.
I mainly work with vision models on mobile and edge hardware, and I'm currently exploring larger models, including LLMs, on resource-constrained devices.
Currently building an on-device AI system with Raspberry Pi 5 + Hailo-8:
- NPU-accelerated vision inference
- A small LLM agent running on the CPU
- Quantization and inference runtime benchmarks
Tools: PyTorch 路 ONNX / ONNX Runtime 路 Hailo DFC / HailoRT 路 ncnn 路 llama.cpp 路 NVIDIA Triton 路 MLflow 路 Docker