Computer Vision Engineer

Role Description

This is a role for someone who wants to do real applied research and ship it. You'll work on a small, highly technical team where the problems are genuinely hard, the resources are there to solve them, and the output ends up in the hands of players, coaches, and front office personnel at some of the most competitive sports programs in the world.

We're open to hiring this role in Eastern Time or Central Time in the US or Canada, and at either the Computer Vision Engineer I or II levels.

• Own the development and deployment of advanced CV models — object detection, multi-object tracking, camera calibration, and semantic segmentation — for player tracking across multiple sports.
• Build and scale high-performance pipelines for video and image data ingestion, processing, and analysis that hold up under the demands of professional and collegiate programs.
• Contribute across the full CV/ML lifecycle, from data gathering, model development and training through production deployment, monitoring, and iterative refinement.
• Drive R&D initiatives including 3D body pose tracking, real-time analysis systems, and integration with LLM/NLP research — with real latitude to explore within the scope of each project.
• Collaborate closely with Software Engineering, Product, and Data Ops to ensure your work is understood, integrated, and used correctly downstream.

Qualifications

• Hands-on proficiency in Python, PyTorch, and OpenCV for production CV applications — not just research prototypes.
• Demonstrated experience taking CV/ML models through the full lifecycle: development, training, evaluation, deployment, and ongoing refinement in real-world systems.
• Ability to design and implement modular, maintainable software architectures for complex CV/ML pipelines — this team builds systems, not just models.
• Cloud computing experience with AWS or similar platforms, plus working knowledge of Linux environments and CUDA for GPU optimization.
• The ability to communicate complex technical work clearly to non-technical partners — when other teams act on your outputs, they need to actually understand what you built and why.

Requirements

• Experience with multi-camera systems, 3D pose estimation, or camera calibration for spatial tracking.
• Background in real-time or near-real-time CV system development with performance optimization.
• Research publication record in CV or ML, open-source contributions, or prior work in sports analytics.

Benefits

• This team has meaningful research latitude — you're not just implementing someone else's architecture.
• Teamworks invests heavily in AI tooling and makes sure engineers have what they need to work at the frontier.
• If you're excited about the intersection of software and sports, there aren't many places doing this work at this scale.