Software Architect – Tracking

This is a fully remote position, open to applicants in United Kingdom.

📋 Description

• Take ownership of software-level architectural design related to tracking, auto-officiating, and data delivery.

• Establish the foundational architecture for new tracking challenges.

• Implement the Software Domain Owner model and maintain designated code owners for shared libraries.

• Ensure design consistency across various sports.

• Represent tracking within the Engineering Architecture Steering group and actively participate in architecture discussions at Hawk-Eye and Sony Sports.

• Determine when to leverage common tracking code and when to take a different approach.

• Lead the direction of computer vision initiatives encompassing camera calibration, ball tracking, skeletal tracking, and related research and development.

• Investigate and introduce innovative tracking technologies.

• Create prototypes to validate proposed ideas.

• Drive the adoption and evaluation of agentic coding frameworks, LLM-driven development tools, and automated code generation within C++ and Python codebases.

• Mentor Squad Tech Leads and provide guidance to junior architects.

• Conduct CV Algorithm talks, HawkTalks, and tutorials.

• Assist in recruitment efforts, including technical assessments for computer vision and machine learning roles.

• Oversee the operational health of ML training and synthetic data infrastructure.

• Identify ownership and support gaps with attached options.

• Serve as the technical escalation point and contribute code when live delivery faces risks.

⛳️ Requirements

• Extensive expertise in multi-camera computer vision.

• Proficiency in camera calibration, including intrinsics, extrinsics, distortion models, anti-wobble techniques, and bundle adjustment.

• Knowledge of triangulation, 3D ball and pose estimation, skeletal tracking, continuous-time modeling, and inverse kinematics.

• Advanced cross-platform C++ programming skills on Windows and Linux, as well as x86 and ARM architectures.

• Familiarity with CMake and real performance profiling techniques.

• Strong understanding of GPU inference, including tools like TensorRT and Triton.

• Solid grasp of ML pipelines, including frameworks such as PyTorch and ONNX.

• Demonstrated experience in running tracking operations at scale in AWS, on-premise environments, and edge devices.

• Practical experience with AI coding agents and LLM orchestration in large codebases.

• Working knowledge of generative AI and synthetic data generation applicable to real-time computer vision.

• Capability to establish architecture across multiple squads and sports while making practical decisions regarding reuse, cost, and time to market.

• Proficient in writing design documents and proposals with options and explanations for non-technical stakeholders.

• Comfortable collaborating with external engineering partners, including Sony, NVIDIA, and AWS, as well as various sports leagues and federations.

• Ability to accurately estimate workload, express confidence levels, and recommend tasks to halt or postpone.

🏝️ Benefits

• Travel opportunities related to the role.

• Chance to collaborate with Sony, NVIDIA, AWS, sports leagues, and federations.

• Access to coaching, mentoring, and professional development through CV Algorithm talks, HawkTalks, and tutorials.