Senior Applied Scientist - Agentic AI (all genders)

Role: Senior Applied Scientist in Customer Care Technology, Berlin. Design, build, and scale Agentic AI systems and Large Language Models (LLMs) for customer self‑service, working with a cross‑functional team of data engineers, software engineers, analysts, and product managers. Responsibilities: - Drive end‑to‑end research by translating high‑impact business challenges into structured research plans and delivering production deployments. - Build, fine‑tune, evaluate, and scale advanced LLM applications and agentic workflows. - Own production models: architecture design, performance optimization, deployment, and real‑time monitoring at scale. - Establish best practices for research methodology and ensure high‑quality, reproducible experiments. - Mentor and coach junior Applied Scientists, fostering scientific rigor and innovation. - Collaborate cross‑functionally with product managers, data engineers, and software engineers to integrate AI solutions into production services. Qualifications: - Proven track record of autonomously tackling complex research projects and shipping AI products at scale. - Deep expertise in Machine Learning, Deep Learning, and Natural Language Processing, with a focus on production‑grade LLMs. - Strong understanding of agentic system design, orchestration, hallucination mitigation, and retrieval‑augmented generation (RAG). - Excellent Python programming skills (including PyTorch, PydanticAI), strong SQL abilities, and hands‑on experience with Databricks. - Ability to translate complex technical concepts into clear insights for non‑experts and senior leadership. - Agile, collaborative mindset comfortable navigating high‑uncertainty environments. What We Offer: - 27–30 days of holiday per year and 2 paid volunteering days. - Employee share programme and discounts (40% off Zalando fashion, 30% off Lounge, external partner offers). - Relocation assistance, family services, health and wellbeing options, and mental health support. - Development opportunities through a training platform and biannual peer‑to‑peer reviews.