Mohammadreza Hosseini
Profile
About
Edge AI / Computer Vision engineer focused on real-time vision systems — from model optimization to production serving on edge devices and Kubernetes.
I’m currently a Computer Vision Engineer at Awentia, where I build C++/GStreamer pipelines on Hailo and NVIDIA stacks (DeepStream, TensorRT, OpenVINO) and deploy models with Triton Inference Server. Recent impact includes reducing manual dataset collection effort by ~40%, cutting deployment time by ~60% through CI/CD, and delivering stable real-time pipelines (e.g. multi-class safety detection at ~30 FPS end-to-end).
In parallel, I’m founding and shipping PassaTeoria (passateoria.it), a cross-platform EdTech product (web, Android, iOS) for the Italian driving theory exam — owning product, full-stack delivery, billing, and cloud operations end to end.
Core stack
- Python
- C++
- PyTorch
- TensorRT
- DeepStream
- GStreamer
- Triton
- Docker
- Kubernetes
- Hailo
- Jetson
- GCP
- CI/CD.
