Computer Vision Engineer
Core
Design, train, and deploy deep learning and classical computer vision models for sports performance analytics, translating product requirements into production-ready artifacts for elite athletes and coaches.
Role type
Full-lifecycle Computer Vision Engineer (Data Science team)
Builds
Production computer vision pipelines, deep learning models, and cloud microservices for sports analytics
Domain
Sports technology / Computer Vision / Deep Learning
Deliverable
production ML models
Required skills
Deep learning architectures, classical computer vision, Python, C++ (reading/debugging), PyTorch or TensorFlow, Docker, AWS, TensorRT or ONNX Runtime, multi-view geometry, object tracking, spatial transformation
Preferred skills
Neural architecture customization, applied linear algebra and matrix calculus, native application development tools (Visual Studio, Qt Creator), sports video benchmarks
Technologies
PyTorch, TensorFlow, OpenCV, Docker, AWS, ONNX Runtime, CMake, TensorRT
Responsibilities
Design and evaluate deep learning architectures alongside classical computer vision pipelines; Develop mathematical pipelines for camera calibration and coordinate mapping; Architect and containerise Python-based cloud microservices; Assist in compiling cross-platform native binaries for desktop applications; Build automated data-ingestion pipelines with model-assisted pre-labeling; Define clean API boundaries for data science modules
Seniority
Mid-Senior, hands-on IC
