Senior Data Scientist - Content and Consumer

About Delivery Hero and the Role

Delivery Hero is a global leader in local delivery platforms, operating in about 65 countries and headquartered in Berlin, Germany. Listed on the Frankfurt Stock Exchange and part of the MDAX, Delivery Hero empowers innovative talent to design impactful solutions. We move swiftly, act decisively, and foster an inclusive environment welcoming diverse backgrounds and perspectives.

We are currently seeking a Senior Data Scientist to join our Content tribe, focusing on building advanced ratings and reviews systems that influence millions of users’ ordering decisions across various markets and languages. This role centers around developing a greenfield system with complex, multilingual user-generated content requiring natural language understanding and machine learning.

Key Responsibilities

• Lead the design and development of LLM (Large Language Model) systems for social proof, ensuring quality, reliability, and comprehensive coverage across languages, platforms, and use cases, including managing system drift, calibration, and data quality in production.
• Establish standards for model and provider selection based on empirical evidence involving cost, latency, and quality metrics, applying systematic prompt engineering and determining the appropriateness of LLMs for specific tasks.
• Drive the product roadmap by identifying high-value problems, quantifying their business impact, and converting them into concrete, manageable projects, treating inference costs as a key product consideration.
• Develop and operationalize meaningful evaluation metrics and infrastructure, such as offline evaluation suites, LLM evaluation frameworks, and annotation workflows, to ensure timely and accurate measurement of generative model performance.
• Collaborate cross-functionally to transition prototypes into production, shaping the system’s architecture and data flows alongside backend and data engineering teams, and creating feedback mechanisms for continual system improvement without manual oversight.
• Mentor peers, promote best practices, and cultivate a pragmatic culture of ownership and high standards within the team.

Essential Qualifications

• Strong expertise in NLP and LLM systems, including experience deploying LLM-based systems on challenging, noisy, multilingual user-generated text, with knowledge of multiple model families and architectures, and the ability to quantitatively assess trade-offs in cost, latency, and quality.
• Proven experience designing rigorous offline evaluation methodologies for generative AI models, including LLM-as-judge frameworks, annotation processes, and measurement of hallucination and fidelity at scale.
• Advanced skills in observability and instrumentation of LLM pipelines to enable debugging, traceability, and cost monitoring, understanding when overhead is justified versus optional.
• Strong product thinking capabilities to translate ambiguous requirements into defined data science problems and identify impactful opportunities linking technical output with business outcomes.
• Comfort with ambiguity and a minimum viable product mindset, capable of delivering incremental value swiftly without overengineering.
• Demonstrated ability to improve team capabilities through mentorship, elevating overall standards and effectiveness beyond individual contributions.

Additional Desirable Skills

• Proficiency in production-ready Python and analytical SQL, familiarity with ML lifecycle management, CI/CD, orchestration, and monitoring tools.
• Solid foundation in statistics and causal inference with a cautious approach to results that may seem overly optimistic.
• Experience with data collection, labeling workflows, and collaboration with annotation teams.
• Familiarity with agentic development tools applied pragmatically, such as automated prompt tuning or LLM-assisted annotation, with the ability to validate outputs effectively.

Additional Information and Benefits

• Hybrid working model with an expectation to be on-site in Berlin campus two days a week for collaboration.
• Generous annual leave entitlements, including 27 days plus an extra day for second and third years of service.
• Support for professional growth including a €1000 educational budget, language courses, parental support, and access to an extensive online learning platform.
• Wellness initiatives including health checkups, meditation, and gym facilities.
• Competitive financial benefits such as employee stock purchase plans, sabbatical options, public transportation discounts, life and accident insurance, and pension schemes.
• Meal benefits including digital and physical food vouchers to encourage team gatherings.
• Relocation support and resources provided for candidates moving to Berlin.
• Commitment to diversity and inclusion including accommodations for disabilities, respectful pronoun usage, and preferential consideration for severely disabled candidates with equivalent qualifications.