Data Scientist - Search Retrieval (f/m/d)

Deine Aufgaben

• Design, build, and improve retrieval passes for our product search, including keyword-based (e.g., BM25) and vector-based (embedding) search
• Develop hybrid retrieval strategies that blend lexical and semantic signals, including candidate merging and score normalization
• Continuously improve retrieval systems by evaluating performance and leveraging interaction data from millions of users
• Analyze search logs and user behavior to identify recall gaps, underperforming query segments, and long-tail query issues
• Bring models from prototype to production together with engineers, ensuring low latency and high availability at e-commerce scale

Deine Fähigkeiten

• University degree in data science or a related field (MSc or PhD), as well as at least 3 years of relevant professional experience, ideally with practical experience in search or information retrieval
• Excellent Python knowledge including experience with machine learning libraries such as PyTorch, TensorFlow and scikit-learn
• Experience designing experiments and interpreting metrics for large-scale online systems (A/B testing, offline evaluation)
• Experience in handling large amounts of data, including advanced SQL knowledge and experience with database systems (e.g. BigQuery) for analyzing user interactions
• Experience in the field of MLOps, especially in the deployment and operation of machine learning models in production, e.g. with orchestration tools like Airflow or Dagster
• Solid command of English, both written and spoken; German is a plus
• Pragmatic, product-oriented mindset and good communication skills for working across disciplines
• A willingness to challenge «how things have always been done» and propose more efficient, automated or effective solutions
• Using creativity and initiative to solve problems and achieve goals

Warum startest du bei uns?