Senior Data Scientist - (Content, Consumer)
On-site · Berlin, State of Berlin, Germany
Job Summary
Senior Data Scientist to own and scale social proof data products end-to-end in Delivery Hero's Content tribe. Build and productionize NLP/LLM solutions for ratings & reviews across multilingual contexts, from problem discovery to model deployment, evaluation, and iterative improvement. Lead the establishment of robust evaluation pipelines, run A/B tests, and drive quantifiable business impact. Collaborate with backend/data engineering on architecture and data flows, write production-grade Python/SQL, contribute to ML lifecycle practices, and mentor teammates. You will work in a fast-moving environment with a hybrid on-site presence in Berlin (2 days/week), focusing on end-to-end data product ownership, LLM decision-making, and building reliable, scalable systems. Optional experience with data labeling pipelines and annotation strategies is a plus. The role emphasizes ownership, pragmatism, and technical excellence, with a strong focus on multilingual NLP, prompt strategies, and credible impact measurement.
Required Qualifications
- NLP & LLM Domain Expertise with hands-on experience in multilingual contexts
- End-to-end ownership of data science solutions from problem definition to production
- Production-grade Python and SQL coding standards; familiarity with ML lifecycle practices
- Experience with building and evaluating ML/LLM systems, including prompt engineering vs fine-tuning decisions
- Establishing evaluation infrastructure, offline evaluation pipelines, annotation processes
- Designing and owning A/B tests and quasi-experiments with credible impact measurement
- Mentoring others and contributing to data science best practices
- Comfort operating in fast-moving, uncertain environments
- Experience with observability, monitoring token usage and costs in LLM pipelines
- Ability to translate ambiguous business problems into well-scoped DS/ML solutions
- Collaborating with backend and data engineering on architecture and data flows
- Optional: experience with data labeling pipelines and annotation strategies
- Strong communication skills and product-minded approach
- Willingness to join hybrid work model at Berlin campus (2 days/week)
- Relocation support for moving to Berlin
Apply with one swipe on Sorce. We auto-fill applications and apply on your behalf — no cover letters, no 40-minute forms.
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.