Senior Computer Vision Engineer - Video Analytics

Description :

About Our Client

Our client is a technology company built on deep expertise in video processing, cloud computing, and embedded systems. They specialize in designing scalable, high-performance solutions to complex, real-world problems - enabling IoT and edge devices to fully leverage cloud and AI capabilities.

Role Overview :

We are seeking a Senior Computer Vision Engineer who thrives at the intersection of modern C++ engineering, classical computer vision, video processing frameworks, and applied machine learning.

Youll drive innovation in real-time video analytics pipelines, optimize ML models for edge and cloud deployment (including on NVIDIA Jetson devices), and guide a growing team of engineers through mentorship and technical leadership.

This is a hands-on technical leadership role - ideal for someone who wants to shape the direction of next-generation video intelligence systems.

Key Responsibilities :

- Lead and own the end-to-end development of computer vision and video analytics modules.

- Design and implement robust, efficient pipelines using modern C++ (C++17/20) and Python.

- Integrate and optimize ML models for inference on CPU, GPU, and Jetson platforms using TensorRT, ONNX, or CUDA.

- Build and tune real-time streaming pipelines using FFmpeg, GStreamer, and related frameworks.

- Develop and enhance classical vision algorithms (edge detection, filtering, object tracking, morphological ops, etc.) alongside modern deep learning solutions.

- Analyze video and image encoding trade-offs across different formats, bitrates, and hardware accelerators.

- Collaborate cross-functionally with Product, QA, and DevOps teams for full-cycle delivery.

- Mentor junior engineers through design reviews, code walkthroughs, and pair programming.

- Drive innovation - explore new frameworks, evaluate hardware capabilities, and improve performance across platforms.

Core Skills :

Programming :

- Expert in C++17/20 (templates, STL, concurrency, performance tuning, RAII).

- Strong scripting and prototyping skills in Python.

- Deep understanding of object-oriented design, modular architecture, and Agile methods.

Computer Vision & Image Processing :

- Solid foundation in OpenCV and related libraries.

- Experience in image enhancement, feature extraction, segmentation, edge/contour detection, and morphological operations.

- Familiarity with color spaces, image formats, and pixel data layouts (YUV, NV12, RGB, grayscale, etc.).

Video Processing & Streaming :

- Experience with FFmpeg, GStreamer, and video codec handling (H.264, H.265, MJPEG, VP9, etc.).

- Understanding of encoding/decoding trade-offs, latency optimization, bitrate management, and hardware acceleration.

- Familiar with streaming protocols like RTSP, RTP, HLS, WebRTC, TCP/IP, UDP, ONVIF.

Machine Learning & Optimization :

- Understanding of ML model inference, optimization, and deployment using TensorRT, ONNX Runtime, or OpenVINO.

- Experience with GPU programming (CUDA/OpenCL) for real-time inference.

- Exposure to model quantization, precision tuning, and hardware-aware optimization.

Development Ecosystem :

- Build tools : CMake, Make, VCPKG

- DevOps : Docker, GitHub Actions, Azure DevOps

- Platforms : Linux (primary), Windows, NVIDIA Jetson (plus)

Preferred / Good-to-Have Experience :

NVIDIA Jetson ecosystem :

- Experience deploying CV and ML pipelines on Jetson Nano / Xavier / Orin.

- Familiarity with JetPack SDK, DeepStream, NvCodec, and hardware-accelerated video pipelines.

- Experience optimizing GStreamer plugins and TensorRT engines for Jetson.

- Understanding of memory bandwidth, power-performance trade-offs, and thermal constraints in embedded vision systems.

- Hardware-specific optimization across CPU, GPU, and embedded accelerators.

- Experience in video analytics, industrial inspection, or surveillance systems.

- Prior experience leading small technical teams or mentoring developers.

- Contributions to open-source projects or research work in computer vision/video analytics.