Job Description
We are on the lookout for a Principal Data Scientist to join the AdTech data science team (Vendor Recommendations and AI) on our journey to always deliver amazing experiences.
As part of our Vendor Team, you’ll be the driving force behind the success of thousands of restaurants, shops, and local businesses. Your contributions will empower vendors with advanced tools to manage their operations, boosting their visibility, and reach. Every feature you help build will create growth opportunities for businesses of all sizes, strengthening Delivery Hero’s ecosystem and impact.
At Delivery Hero, AdTech introduces people to new food they love and enables our partners to meet their future customers at the perfect moment. We have developed a range of products to offer advertising solutions for restaurants helping them to increase their visibility and reach, improve their order conversion, and eventually drive more sales by covering the entire marketing funnel. Operating across over 50 countries around the clock, our Ad Tech connects the ideal advertiser with the ideal customer millions of times every day. Ad Tech is a fundamental area for us on our path to profitability. our data science team is at the heart of this growth, powering personalisation, Ranking, relevance, and recommendation.
1. End-to-End Ownership: Lead the design, build, and deployment of large-scale AI recommendation systems that deliver multi-million-euro business outcomes for the vendor vertical.
2. Next Best Action (NBA) Framework: Own the architecture and modeling for determining the optimal sequence and timing of actions (eg ad suggestions, promotional offers, operational advice) to maximize vendor success metrics.
3. Strategic Time-Series & Forecasting: Lead all core initiatives in advanced time series modeling for demand prediction, vendor forecasting, and understanding long-term trends in the vendor base.
4. Commercial Insights & Take Rate: Act as the technical authority for take rate optimisation, using statistical rigour and causal methods to measure the incrementality and efficiency of all vendor-facing pricing and adtech levers.
5. Cross-Functional Alignment: Act as the primary technical partner for Product, Engineering, and Commercial teams to validate or debunk business hypotheses and drive alignment on the recommendations strategy.
Qualifications
Recommendation Systems & Sequential Authority:
6. Recommendation Architect: Expert practitioner in designing, building, and deploying large-scale, high-performance recommendation systems (, two-tower, deep sequence models, Graph Neural Networks) for the B2B or marketplace domain.
7. Sequential/Timeseries Specialist: Thought leader in using timeseries analysis and sequence modeling (eg LSTMs, transformers, advanced forecasting) for vendor behavior, next-best-action prediction, and demand forecasting.
8. Causal Inference for Action: Proven ability to apply causal inference and advanced experimentation frameworks to quantify the true incremental impact of recommendations and vendor-facing product features.
9. Take Rate Analysis: Deep expertise in the analytical frameworks and models required for optimizing key commercial metrics like take rate, vendor LTV, and churn prediction.
Advanced AI & Statistical Engineering:
10. Multi-Objective Optimization: Mastery in formulating and solving problems that balance potentially conflicting vendor KPIs - such as maximizing short-term revenue/take rate while maintaining vendor engagement and long-term retention/success.
11. Algorithmic Decisioning: Deep experience in designing and deploying real-time decision-making algorithms that power "Next Best Action" systems and adapt to shifting vendor and market conditions.
12. Probabilistic Modeling: Strong foundation in Bayesian methods and uncertainty quantification to provide reliable forecasts and confidence intervals around business-critical projections.
13. Production Excellence: Deep operational knowledge of deploying these sophisticated models using Python frameworks like Keras, scikit-learn.
Strategic Leadership & Influence:
14. The "Recommendation North Star": Proven ability to define a 2+ year technical roadmap for Vendor Recommendation Data Science, influencing the strategic direction of Product, Engineering, and Commercial leadership.
15. Decision Science Mastery: A track record of translating complex ML/statistical findings into clear, actionable executive narratives (eg explaining the trade-offs of a new recommendation model to senior leadership).
16. Force Multiplier: A history of mentoring Staff and Senior Data Scientists, raising the bar for scientific rigour and production ML excellence across the entire organisation.
Nice-to-Haves
17. Expertise in Reinforcement Learning (RL), particularly in applying Multi-Agent RL or contextual bandits to sequential decision-making and next-best-action problems.
18. Familiarity with Game Theory or Mechanism Design as applied to multi-sided platform optimisation.
19. Major Contributions (papers, talks, or libraries) in the fields of Large-Scale Recommendation Systems or Time-Series Forecasting.
20. Expertise in designing feature stores and serving pipelines for sequence and time-series features.