Senior Data Scientist, Growth Alliance

HelloFresh is a global company with over 1000 people in its HelloTech organization, building digital products for end-to-end food experiences including meal kits, ready-to-eat meals, pet food, and premium meat & seafood. HelloTech operates with autonomous, cross-functional alliances responsible for specific products or domains, supporting brands with shared platforms, data, and operational excellence. The role is based in Warsaw at HelloFresh's Warsaw office, in the Communications tribe within the Growth Alliance.

• 5+ years of professional experience in data science, ML, or related quantitative field with a track record of deploying models to production.
• Proficient in Python and experience with modern ML stack tools such as Databricks/Spark and MLflow.
• Strong foundation in machine learning for personalization, with direct experience in reinforcement learning and/or contextual bandits (e.g., Thompson Sampling, epsilon-greedy).
• Solid knowledge of experimentation and causal inference, designing and analyzing controlled experiments in high-traffic settings.
• Experience with classical ML techniques including regression, classification, clustering, and recommendation systems.
• Experience handling customer behavior data, segmentation, and production feature pipelines.
• Clear, structured communication and collaborative, cross-functional mindset.

Nice to have:

• Off-policy evaluation/counterfactual learning, propensity methods, or sequential testing (SPRT).
• Experience with experimentation platforms like Statsig and streaming/low-latency inference.
• Experience with unstructured data features such as embeddings and text signals.
• Exposure to marketing/CRM decisioning, deliverability, or funnel/conversion optimization.

• Build, own, and iterate on reinforcement learning (RL) and machine learning (ML) models in production behind high-traffic conversion and communication flows.
• Develop decisioning logic to determine the best next action for each user across the funnel and messaging lifecycle.
• Drive architectural evolution of contextual-bandit systems from online exploration to offline policy evaluation and low-latency inference.
• Partner with Engineering on system design for clean integration of models into funnel and messaging architecture.
• Own experimentation design for decisioning models including success metrics, guardrails, and result interpretation using rigorous methods.
• Monitor model performance, input data reliability, drift, and latency, iterating quickly when issues arise.
• Collaborate with CRM Ops, Marketing Analytics, and platform partners to translate model behavior into trusted marketing decisions.