Ort Germany - Berlin Vertrag Full time Job-Kategorie Applied Science & Research

Applied Scientist (all genders) - B2B Data Science and Algorithms

The ZEOS department is responsible for all partner-facing Zalando Logistics Solutions. We provide a holistic approach to delivering the fulfillment solutions that meet our partner’s needs by unifying these services under a single umbrella. We aim to provide our partners with a profitable fulfilment experience, and we see that to do this, Machine Learning, Operations Research and Data-driven solutions will play a pivotal role.

We are seeking an Applied Scientist who is fueled by the desire to build innovative and impactful ML/Optimization systems for our B2B logistics partners. You will join an existing team of Applied Scientists, and Machine Learning Engineers and work in a cross-functional setup with Product Managers, Data and Software Engineers.

The team’s focus is helping our B2B partners improve inventory health and order fulfilment efficiency. You will be building various ML/DL forecasting models (demand, returns, lead-times), stochastic inventory optimization solutions, recommendation services and emerging Agentic AI systems that generate these key insights for our partners and enable them to make data-driven decisions about their article assortment and inventory. You will play a key role in a cross-functional team!

WHY YOU SHOULD BE INTERESTED…

• Shape the Product from Day One: You won't be a "model factory" stuck in a silo! You will be a core scientific partner, working directly with product and user research to define the problem, not just solve it. Your expertise will directly influence the product roadmap and identify new opportunities to create value for our partners.

• Solve High-Stakes, Complex Scientific Problems: This is your chance to go beyond standard forecasting. You'll be tackling cutting-edge challenges in stochastic inventory optimization, multi-echelon demand forecasting (including returns and lead-times), and building recommender systems for assortment planning. Your work will be at the absolute core of our partners' profitability.

• Pioneer Agentic AI & Robust Evaluations: As we explore opening up MCP servers to our partners, you will contribute to defining how we measure their performance/success in the user’s agentic journey. You will design and implement rigorous Agentic AI evaluation frameworks/metrics (e.g., LLM-as-a-judge, trajectory evaluation, and safety guardrails) to ensure the autonomous systems connecting to the MCPs act reliably and logically on behalf of our B2B partners.

• End-to-End Ownership: You will have the autonomy to see your ideas through from initial research and prototyping all the way to production. You will define the metrics for success, collaborate with engineers to deploy your models as scalable services, and monitor their real-world impact on partner KPIs.

• Drive Impact at a multi-merchant scale: Your solutions won't just help one partner. You will be building platform-level services that scale across hundreds of diverse partners/merchants, directly influencing millions of euros in merchandise value and shaping the future of a more sustainable and efficient e-commerce logistics network.

WE’D LOVE TO MEET YOU IF…

• Educational background in a quantitative field - Masters degree or higher preferred.

• 3+ years of hands-on industry experience in an Applied Scientist, Data Scientist, or Research Scientist role, applying scientific methods to solve business problems.

• Industry demonstrated knowledge and skills in at least one the following areas:

  • Machine Learning or Deep Learning, particularly applied for time-series forecasting (e.g., LGBM, ARIMA, Prophet, Transformers, Nixtla, Darts,...)
  • Machine Learning Engineering (e.g. service design, batch processing, G