<div class="show-more-less-html__markup show-more-less-html__markup--clamp-after-5 relative overflow-hidden"> <p>Headquartered in North Germany, one of Germany's largest energy service providers is doubling its investment into AI. Renewable energy output is unpredictable and the trading desk needs to be smarter to manage that volatility and they are building the ML/AI capability to do it.</p><p><br/></p><p><strong>The Role </strong></p><p><br/></p><ul><li>Senior ML Engineer focused on hands-on development, building energy trading ML models from the ground up.</li><li>Design and deploy autonomous trading systems, time series forecasting models, and supervised/unsupervised ML algorithms specifically for power and gas markets.</li><li>Own the end-to-end lifecycle: from back-testing and strategy optimisation to deploying prediction models into production.</li><li>Architect MLOps pipelines and live data infrastructure that feed directly into real-time trading decisions.</li><li>Take technical ownership of the modelling roadmap. This is a high-impact IC role with no direct reports, but with significant influence on trading strategy.</li></ul><p><br/></p><p><strong>The Candidate</strong></p><p><br/></p><ul><li>Deep mathematical and statistical background with extensive experience in time series analysis, classical ML (Random Forest, Gradient Boosting), and deep learning for sequential data.</li><li>Proven track record building automated or algorithmic trading systems—specifically in energy markets (power/gas) or high-frequency financial markets.</li><li>Expert level Python skills (NumPy, Pandas, Scikit-learn, PyTorch/TensorFlow) and experience with production-grade coding (version control, testing, CI/CD).</li><li>Strong understanding of MLOps (Docker, Kubernetes, feature stores, model monitoring) and real-time data pipelines (Kafka, Redis, or similar).</li><li>Fluent in English (working language); German is a bonus but not required.</li><li>95% remote with occasional travel to North Germany for team sprints and strategy meetings.</li></ul><p></p> </div>