Data Science Manager

Visa Visa · Fintech · Frankfurt, Germany, DE

Visa is seeking a Data Science Manager to co-develop data-driven solutions for clients using advanced analytics and AI techniques. The role involves understanding Visa's data assets, partnering with sales and product teams, defining business problems, executing data science projects using statistical and AI methods, and feeding insights into product development. The ideal candidate will have experience with predictive modeling, machine learning, experimentation, agentic AI systems, data preparation, Python, SQL, and LLM integration patterns like RAG and tool calling.

What you'd actually do

  1. Develop a deep understanding of Visa’s data assets and value‑added services, and proactively translate them into compelling, commercially viable propositions that address client needs across acquisition, usage, and retention.
  2. Partnering closely with Sales, VCA, and Product teams to identify opportunities, shape client conversations, and convert insights into funded projects.
  3. Collaborate with internal and external stakeholders to define clear business problems, commercial outcomes, and success metrics, translating them into structured analytical and delivery plans.
  4. Execute the Data Science projects, ensuring the use of appropriate statistical and AI techniques to generate clear, business‑centric insights, supported by strong storytelling and impactful visualisation.
  5. Provide subject‑matter expertise and quality assurance across complex Data Science engagements, ensuring analytical rigor, relevance, and alignment with client objectives.

Skills

Required

  • Advanced analytics and AI experience applying a range of techniques (e.g., predictive modeling, machine learning, experimentation, agentic AI systems) to solve real business problems and drive measurable outcomes.
  • Strong hands-on capability in data preparation and feature engineering, including cleaning, transforming, and validating large, complex datasets.
  • Programming skills in Python and SQL, including production-grade data manipulation and ML workflows (e.g., pandas, scikit-learn or equivalent libraries), and the ability to work efficiently with large datasets.
  • Experience designing and implementing agentic AI workflows, including multi-step reasoning pipelines, tool-augmented agents, and orchestration frameworks (e.g., LangChain, LangGraph, AutoGen, or equivalent), with the ability to evaluate agent reliability and manage failure modes in production.
  • Familiarity with large language model (LLM) integration patterns, including prompt engineering, retrieval-augmented generation (RAG), function/tool calling, and memory management within agentic architectures.
  • Proven ability to translate business needs into end-to-end analytical/AI solutions, from problem framing and methodology design to insight delivery and stakeholder adoption — including solutions that leverage autonomous or semi-autonomous AI agents.
  • Ability to communicate complex technical concepts clearly and credibly to non-technical stakeholders, influencing decisions through storytelling, visualization, and structured recommendations.

What the JD emphasized

  • agentic AI systems
  • agentic AI workflows
  • agent reliability
  • manage failure modes in production
  • autonomous or semi-autonomous AI agents

Other signals

  • AI/ML techniques
  • agentic AI systems
  • agentic AI workflows
  • LLM integration