Design and deploy end-to-end computer vision pipelines using NVIDIA DeepStream and Jetson platforms
Develop and fine-tune object detection, tracking, and action recognition models using PyTorch and YOLO
Build and optimize real-time video streaming and inference solutions
Integrate vision systems with NVIDIA ISAAC for robotics or spatial AI use cases
Collaborate with product and engineering teams to ship features from prototype to production
Profile and optimize model performance for low-latency edge deployment
Computer Vision Engineer
Islamabad
Full Time
Onsite
Posted
Key Responsibilities
Skills and Qualifications
2–4 years of hands-on experience in computer vision or related fields
Proficiency with PyTorch and YOLO-based detection frameworks (YOLOv8/v9/v10)
Practical experience with NVIDIA DeepStream SDK and Jetson devices (Nano, Orin, AGX)
Familiarity with action recognition architectures (SlowFast, I3D, or similar)
Experience building real-time streaming pipelines (GStreamer, RTSP, WebRTC)
Strong understanding of model optimization techniques (TensorRT, quantization, pruning)
Solid Python and C++ skills
Bonus Points
Experience with NVIDIA ISAAC SDK or Isaac ROS
Background in healthcare or surveillance AI applications
Familiarity with MLOps and model deployment workflows
What We Offer
Work on high-impact AI systems used in real deployments
Collaborative, fast-moving team with a strong engineering culture
Competitive salary and growth opportunities
Be part of a company incubated at National Science & Technology Park, Islamabad