Machine Learning Engineer (m/f/d)

We are looking for a Machine Learning Engineer to join Sportradar's AI unit, an established, high performing team of AI experts working on the design, development and deployment of AI models that deepen our understanding of sports and enable new sports technology products for sports performance, betting and media. The position will focus on the research and development of generative models of sports gameplay and predictive models of player actions and performance. You will collaborate closely with product owners and technical leads to ensure these models translate into real product impact across diverse areas such as coaching, sports betting, augmented streaming and sports virtualization.

• Analyse, explore and visualize datasets using statistical analysis and scripting, including analysis and preparation of large-scale body pose tracking data collections.
• Develop and train machine learning models, specializing in deep learning methods for sequence modelling, including autoregressive or diffusion-based generative models.
• Design, train, and iterate on representation learning solutions optimized for modelling of sports data with LLM-like models.
• Develop data processing pipelines adopted for training models from scratch, validation, and inferencing.
• Rigorously validate methods, models, algorithms, and hypotheses using back-testing on historical data and/or simulations.
• Optimize models for efficient GPU-accelerated inferencing at scale.
• Bring models into production leveraging internal AI platform, maintaining ownership of the AI solution throughout the model development lifecycle, from modelling to production deployment, monitoring and continuous model improvement.
• Collaborate closely with stakeholders to connect model outputs to product experiences (e.g., real-time visualisation, coaching workflows, simulated reality), balancing research progress with practical delivery constraints.
• Present ideas and solutions to software developers and business stakeholders in a clear and understandable way.

• Knowledgeable in modern machine learning techniques with strong fundamentals in deep learning for sequence modeling and generative AI (required).
• Strong programming skills and software development experience to build reliable training pipelines, scalable data processing, and production-grade inference services (required).
• Hands-on experience working with spatiotemporal data (tracking/trajectory data, time-series, or sensor-like modalities) and building models that learn from complex structured inputs (preferred).
• Proficiency with Python and common ML tooling, like PyTorch and PyTorch Lightning, including distributed training and experiment tracking (required).
• Solid understanding of data engineering basics (efficient dataset construction, feature pipelines, validation, reproducibility) (required).
• Experience working with cloud services (AWS) and container-based development (using Docker, Singularity, etc.) (required).
• Ability to properly use JIRA, plan and structure work in efficient and clear manner (required).
• Experience with computer vision / pose estimation pipelines or working directly with pose-tracking outputs and skeleton-based modeling (nice-to-have).
• Bachelor of Science in Computer Science / Engineering, Mathematics / Statistics, or related field; equivalent experience acceptable (required).
• Must be comfortable with using AI coding tools and agentic technologies (required).
• Fluent in English (written and spoken) (required).

• A collaborative environment with colleagues from all over the world (Engineering offices in Europe, Asia and US) including various social events and teambuilding.
• Flexibility to manage your workday and tasks with autonomy.
• A balance of structure and autonomy to tackle your daily tasks.
• Vibrant and inclusive community, including Women in Tech and Pride groups which welcome all participants.
• Global Employee Assistance Programme.
• Calm and Reulay app (leading well-being apps designed to support focus, quality rest, mindfulness, and long-term mental resilience).
• Online training videos.
• Flexible working hours.