Muhammad A.

I build production Computer Vision and Edge AI systems for real-time video, multi-camera analytics, object detection and tracking, NVIDIA Jetson deployment, and performance-critical AI applications. $600K+ earned on Upwork | 13,600+ hours | 137 jobs | 100% Job Success | Top Rated Plus My work covers the complete Computer Vision pipeline, from model development and optimization to camera integration, edge deployment, cloud connectivity, and production monitoring. Recent work includes: • $100K+ object detection and counting system • $31K+ Jetson / ARM / TensorRT development • $23K+ Jetson Computer Vision security pipeline • $23K+ Computer Vision pipeline analysis and optimization • $16K image feature detection system • $13K ROS2 LiDAR fusion and multi sensor tracking project • $10K+ NVR pipeline on NVIDIA Jetson Orin • $10K+ vehicle detection AI system My core expertise includes: • Computer Vision, object detection, tracking and segmentation • NVIDIA Jetson Nano, Xavier and Orin • DeepStream, CUDA, TensorRT and model optimization • RTSP and multi-camera video analytics • ROS2, LiDAR and multi sensor perception • OpenCV, PyTorch, TensorFlow, YOLO, Python and C++ • GStreamer, Linux, Docker, AWS and GCP • OCR, pose estimation, gesture recognition and image processing I regularly help clients solve problems such as low FPS, high latency, unstable camera pipelines, poor detection accuracy, multi-camera scaling, model deployment, and edge hardware optimization. I also support LLM, RAG and AI agent systems when conversational AI is part of the broader product. If you’ve made it this far, you probably have a Computer Vision or Edge AI problem worth solving. Send me a message or invite me to your job and tell me what you’re building, what hardware you’re using, and where the bottleneck is. I’ll take a look and suggest the best way forward. If the fit looks right, we can get started right away.