NVIDIA DeepStream SDK Expert

We are seeking a highly skilled Senior Software Engineer with deep expertise in NVIDIA DeepStream SDK to join our R&D team. In this role, you will design, develop, and optimize high-performance video analytics pipelines for real-time applications using DeepStream, GStreamer, and NVIDIA's GPU-accelerated frameworks. You will work closely with ML engineers, system architects, and product teams to bring cutting-edge video intelligence solutions to production.

Key Responsibilities

• Design and implement scalable video analytics pipelines using NVIDIA DeepStream, GStreamer, and related NVIDIA SDKs.
• Develop and optimize custom DeepStream plugins and modules for video decoding, inference, tracking, and visualization.
• Integrate AI models (e.g., YOLO, SSD, ResNet) into DeepStream pipelines for real-time inference.
• Work with Jetson and GPU platforms to deliver low-latency, high-throughput solutions for edge and cloud deployments.
• Collaborate with cross-functional teams on performance tuning, deployment, and monitoring of computer vision systems.
• Contribute to system architecture decisions, benchmarking strategies, and code reviews.
• Document pipeline components, APIs, and configuration strategies for maintainability and knowledge transfer.

Must-Have:

• 8+ years of software engineering experience, with at least 5 years working with NVIDIA DeepStream SDK.
• Strong proficiency in C/C++, Python, and GStreamer plugin development.
• Hands-on experience with CUDA, TensorRT, and NVIDIA Jetson platforms.
• Deep understanding of video encoding/decoding, computer vision workflows, and AI inference.
• Experience deploying and optimizing AI/ML models in production environments.

Nice-to-Have:

• Experience with Docker, Kubernetes, or cloud-native deployment strategies.
• Familiarity with RTSP, RTP, WebRTC, or other streaming protocols.
• Background in OpenCV, ROS, or other robotics/video processing libraries.
• Prior work with real-time surveillance, smart city, retail analytics, or autonomous systems.