We’re the world’s leading sports technology company, at the intersection between sports, media, and betting. More than 1,700 sports federations, media outlets, betting operators, and consumer platforms across 120 countries rely on our know-how and technology to boost their business.
ABOUT US:
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.
Sportradar is a global leader in understanding and leveraging the power of sports data for hundreds of business customers around the world, in turn entertaining millions of sports fans.
THE CHALLENGE:
- 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 (not only fine-tuning), 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.
YOUR PROFILE:
- The successful candidate will be knowledgeable in modern machine learning techniques with strong fundamentals in deep learning for sequence modeling and generative AI (e.g., transformers, autoregressive modeling, representation learning)
- Strong programming skills and software development experience to build reliable training pipelines, scalable data processing, and production-grade inference services are a must.
- 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 is highly valued.
- Proficiency with Python and common ML tooling, like PyTorch and PyTorch Lightning, including distributed training and experiment tracking.
- Solid understanding of data engineering basics (efficient dataset construction, feature pipelines, validation, reproducibility).
- Experience working with cloud services (AWS) and container-based development (using Docker, Singularity, etc.)
- Ability to properly use JIRA, plan and structure work in efficient and clear manner.
- Bonus: Experience with computer vision / pose estimation pipelines or working directly with pose-tracking outputs and skeleton-based modeling.
- Bachelor of Science in Computer Science / Engineering, Mathematics / Statistics, or related field; equivalent experience acceptable.
- Must be comfortable with using AI coding tools and agentic technologies.
- Fluent in English (written and spoken).
- Autonomous, rigorous, creative and a team player.
OUR OFFER:
- 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.
OUR RECRUITMENT PROCESS:
- Initial Screening: A quick chat with our Talent Acquisition Partner to understand your background and expectations.
- Technical Assessment: A short (home) task to showcase your technical skills.
- Technical Interview: Meet with the Technical team and later with the Hiring Manager to dive into your solution, as also discuss team fit.
- Onsite Interview: Meet with the local team and take a tour of our office for a final meet-and-greet.
- Finals Steps: Receive feedback and, if successful, an offer!
At Sportradar, we celebrate our diverse group of hardworking employees. Sportradar is committed to ensuring equal access to its programs, facilities, and employment opportunities. All qualified applicants will receive consideration for employment without regard to age, race, color, religion, sex, sexual orientation, gender identity, national origin, disability, or status as a protected veteran. We encourage you to apply even if you only meet most of the requirements (but not 100% of the listed criteria) – we believe skills evolve over time. If you’re willing to learn and grow with us, we invite you to join our team!
