Senior Data Scientist

Senior Data Scientist position in the Machine Learning & Artificial Intelligence unit within Bayer’s Enterprise Data & Analytics Platform. The role involves delivering high-impact AI solutions across Finance, Supply Chain, HR, Procurement, Legal, and Communications. The international team spans Poland, Germany, Spain, and India, working with LLMs, embeddings, classical ML, and optimization on a modern, cloud-native stack (Python, AWS, Azure, Databricks).

• Master’s or PhD with 5+ years in Data Science or Applied ML, delivering production-impact solutions.
• Strong Python and SQL skills; expertise with pandas, NumPy, scikit-learn; experience with PyTorch or TensorFlow is a plus.
• Hands-on with Generative AI: embeddings, prompt engineering, tool/function calling, and agent frameworks (LangChain, LangGraph, PydanticAI); experience with vector databases (pgvector and others).
• Solid grounding in classical ML: model selection, validation, and metrics (e.g., AUC/F1/RMSE); time series or forecasting experience is a plus.
• Evaluation focus: design and run offline/online tests, rubric-based GenAI evaluation, safety checks, and error analysis; familiarity with LangSmith/Langfuse or similar is beneficial.
• Clear stakeholder communication: requirements gathering, expectation setting, storytelling, and influencing decisions with data.
• Proven problem-solving skills: structuring ambiguous problems, hypothesis-driven analysis, and iterative experimentation.
• Basic cloud proficiency (AWS and/or Azure): storage and compute (e.g., S3/Blob, Lambda/Functions or containers), secrets, and Databricks or Spark; awareness of CI/CD (e.g., GitHub Actions).
• Good engineering hygiene: modular code, testing, documentation, and reproducibility.
• Data governance and privacy awareness.
• Fluent in English (written and spoken); additional languages from team regions are a plus.

• Translate business needs into data science problems; define hypotheses, success metrics, and evaluation plans; communicate progress and outcomes to stakeholders.
• Design, build, and evaluate (Gen)AI solutions for enterprise data, including LLMs and AgenticAI; prompt/chain design, tool use, and lightweight agent workflows (LangChain, LangGraph, PydanticAI).
• Develop classical ML models (classification, regression, forecasting, clustering), feature pipelines, and explainability (e.g., SHAP, RAG pipelines and vector search for knowledge use cases).
• Establish rigorous evaluation: offline metrics, human-in-the-loop reviews, rubric-based GenAI assessments, groundedness/hallucination checks, and online tests (A/B or interleaving).
• Implement production-ready Python code with high-quality engineering standards: Git-based workflows, code reviews, automated tests, documentation, and reproducibility.
• Ensure monitoring and traceability: data/feature drift checks, model and prompt versioning, cost/latency tracking, and structured logging.
• Partner cross-functionally with Product, Data Engineers, AI Engineers, and Business stakeholders to drive adoption for high outcome, production-grade AI solutions.
• Champion problem-solving and analytical rigor; lead workshops, stakeholder demos, and clear storytelling around insights, risks, and trade-offs.