Computer Vision Engineer

Catapult is building the future of sports performance technology, with a mission to Unleash the Potential of every athlete and team on earth. We don't just work in the sporting industry; we are actively changing it. Since 2006, our solutions have been leading the way in sports performance software, science, and data, in a world where 1% can literally mean the difference between winning and losing.

We work with over 5,000+ teams around the world, empowering coaches, managers and trainers in premier teams in the NFL, NBA, NHL, MLS, EPL, AFL, NRL, NCAA and more. We provide the information they need to optimise athletes' health, game-day readiness, and performance, as well as in-game tactics. Catapult is a sports technology company that empowers professional teams to make data-driven decisions.

We deliver health, performance, video, and AI insights from the locker room to competitive environments, ensuring every decision is an opportunity to gain an advantage, sharpen performance, and build lasting success. WE WANT PEOPLE WHO ARE PASSIONATE ABOUT DELIVERING INNOVATIVE SOLUTIONS TO COMPLEX PROBLEMS

• End-to-End Pipeline Contribution:
  Collaborate with senior data scientists, computer vision engineers and vertical teams to translate product requirements into practical computer vision solutions, helping design the pipeline from raw video ingestion to production inference.
• Algorithm & Model Development :
  Design, train, and evaluate deep learning architectures alongside classical computer vision pipelines (e.g., feature tracking, optical flow, and spatial filtering via OpenCV).
• Geometric Computer Vision :
  Develop robust mathematical pipelines for camera calibration, homography estimation, and coordinate mapping to ensure model spatial outputs are accurate and stable.
• Modern Cloud & Containerised Deployment:
  Focus on architecting and containerising Python-based cloud microservices (via Docker) as our primary, future-facing deployment model.
• Desktop Applications Support:
  Assist in compiling cross-platform native binaries or shared libraries linking against the ONNX Runtime C++ API to support and maintain our existing Windows/macOS desktop application footprint.
• Automated Data Curation :
  Help build intelligent, automated data-ingestion pipelines that utilise model-assisted pre-labeling to continuously clean and version high-throughput training datasets.
• Interface & Boundary Design:
  Participate in defining clean API boundaries and interface contracts to ensure our core data science modules integrate seamlessly into downstream vertical applications.

• Core Algorithmic Background:
  Foundational knowledge of classical computer vision (multi-view geometry, object tracking, spatial transformation) and modern deep learning architectures (object detection, semantic/instance segmentation, transformer-based vision models).
• Production Model Training:
  Experience sourcing, structuring, training, and benchmarking deep neural networks using PyTorch or Tensor Flow.
• Hybrid Language
  
  Skills:
  
  High proficiency in Python for prototyping, training, scripting, and deployment pipelines, combined with a practical, supporting capability to read, build, and debug existing C++ codebases (including exposure to build management tools like CMake).
• Execution Graph Optimisation:
  Familiarity with optimising model runtimes and inference execution graphs for real-time applications using TensorRT or ONNX Runtime (e.g., quantisation, layer fusion).
• Modern Infrastructure:
  Practical experience with Docker containerisation, version control (Git), and cloud platform execution (AWS).
• NICE TO HAVE Neural Architecture Customisation:
  Experience modifying, adapting, or designing custom neural network components (e.g., specialised backbones, attention mechanisms, or custom loss functions) rather than just implementing standard off-the-shelf models.
• Advanced Mathematical Foundations: A strong intuitive grasp or academic background in applied linear algebra and matrix calculus, particularly as it relates to 3D spatial transformations and projective geometry.
• Downstream Integration:
  Experience or familiarity with native application development tools (Visual Studio, Qt Creator) to help ease collaboration when handing off components to vertical app teams.
• Sports Video Benchmarks:
  Experience experimenting with or competing in open-source sports analytics datasets and challenges (e.g., Soccer Net, SportsMOT, or similar multi-object tracking and action-spotting benchmarks).
• Domain Alignment: A genuine interest in sports analytics, tracking technology, or elite human performance.

In 6 Months' Time... Global Impact:
Your work will be actively contributing to features underpinning informed decisions made by elite coaches and professional athletes globally. Collaborative Innovation:
You will have partnered with the team to take an algorithmic feature from an abstract brief to a stable, deployable Python asset - actively bringing your own…