Role Overview

A data-driven technology company is looking to hire a Lead Computer Vision Engineer to join a growing data function. This is a high-impact role where you will take ownership of designing and scaling systems that extract structured, high-value data from large-scale, complex video sources. Sitting at the intersection of computer vision, data engineering, and applied analytics, you will act as the most senior CV specialist in the organisation, setting technical direction and moving an existing semi-manual process towards a highly automated, reliable pipeline.

Key Responsibilities

• End-to-End Pipeline Development: Design, build, and deploy robust computer vision pipelines to extract structured events and signals from long-form, variable-quality video footage (both historical and near-real-time).
• System Scaling & Automation: Evolve and scale processing systems across large archives of historical footage, identifying and implementing opportunities to reduce manual intervention and increase automation.
• Validation & Quality Assurance: Create comprehensive validation frameworks, metrics, and monitoring to ensure the accuracy and reliability of model outputs, collaborating with domain specialists to define rules and handle edge cases.
• Technical Leadership & Strategy: Act as the subject matter expert for computer vision, owning architectural decisions, defining technical standards, and guiding the long-term strategy for the CV function.

Required Skills & Experience

• Proven commercial experience building and deploying computer vision systems in production environments.
• Strong practical knowledge in one or more of the following: object detection, tracking, temporal event detection, pose estimation, or action recognition.
• Proficiency in Python and common CV/ML libraries (e.g., OpenCV, PyTorch, TensorFlow).
• Experience owning technical and architectural decisions for data-intensive systems.
• A pragmatic mindset, with the ability to balance experimentation with the need to deliver robust, scalable solutions.

Nice-to-Have

• Experience working with noisy, imperfect real-world video data (e.g., broadcast footage, user-generated content).
• Exposure to MLOps principles and tools for deploying and monitoring models at scale.
• Familiarity with sports, media, or broadcast video analytics domains.