Computer Vision Engineer

Muhammad Hassan Gul designs and ships real-time computer vision systems — pose estimation, object tracking and sports analytics — that run in under 30 milliseconds on edge devices. Based in Islamabad, Pakistan, and open to remote or relocation roles worldwide.

Pose Estimation Object Tracking Edge AI PyTorch YOLO

What is a Computer Vision Engineer?

A Computer Vision Engineer builds software that gives machines the ability to see and interpret visual data — images, video and live streams. The work sits at the intersection of deep learning, classical image processing and production engineering: training models, optimizing inference, and deploying them so they run reliably in the real world.

Hassan brings this full stack to every engagement, from dataset design and model training to edge and cloud deployment. At ID Sports Ventures in Berlin, he engineers real-time video analytics that track athlete performance during live competition — turning raw footage into biomechanical insight used by sports scientists.

<30msEdge inference latency
+15%Pose accuracy boost
−40%Pipeline overhead cut

Core Computer Vision Services

Pose estimation & keypoint detection

Lightweight keypoint models that localize joints and limbs from video in real time. Used for biomechanical analysis, form assessment and automated scoring in sports technology. See the dedicated pose estimation engineer page.

Object detection & multi-object tracking

YOLO-based detection combined with ByteTrack to keep consistent identities on every player through severe occlusion — the backbone of automated game analysis and drill assessment.

Real-time video analytics

End-to-end pipelines that ingest RTSP and live streams, process frames at speed, and emit metrics for dashboards. Optimized with ONNX Runtime, TensorRT and CUDA to hold latency targets on consumer-grade hardware.

Why Teams Hire Hassan

  • Sub-30ms inference — models quantized and TensorRT-accelerated for real-time edge deployment on streaming footage.
  • Sports analytics depth — multi-camera pose data translated into standardized biomechanical performance reports.
  • Production MLOps — FastAPI backends, Docker containers and optimized cloud inference (AWS S3, EC2).
  • Clean engineering — 85% test coverage and measurable results: a 15% accuracy boost and 40% less preprocessing overhead.
  • Full-stack vision — Python, C++, OpenCV, PyTorch, TensorFlow and deployment tooling across the board.

Tools & Technologies

The frameworks Hassan works with daily: PyTorch, TensorFlow, OpenCV, YOLO, ByteTrack, DeepSORT, ONNX Runtime, TensorRT, FastAPI, Docker, CUDA, C++ and AWS. He pairs these with optical flow, action recognition and biomechanical metric extraction for motion-focused projects.

How to Hire a Computer Vision Engineer

Reach out with your project goals — latency targets, data sources, team structure — and get a straight answer on feasibility. Hassan is available for full-time roles, contract work and consulting across sports technology, manufacturing, healthcare and robotics applications.

“I engineer computer vision systems that read human motion at machine speed — from sub-30ms edge inference to production-grade deep learning pipelines.” — Muhammad Hassan Gul

Need a Computer Vision Engineer who ships?

Tell me about your latency targets and data — I'll give you a straight answer on feasibility within 48 hours.

Email Hassan