Чтобы адаптировать резюме под вакансию или составить сопроводительное письмо, загрузи резюме
описание
Deliveroo is part of DoorDash and operates a global, three-sided marketplace serving customers, riders, and partners. Its Global Engineering teams build and scale systems for millions of people, while its Consumer organization helps customers discover relevant content.
задачи
Design, develop, and improve recommendation, ranking, and retrieval models that surface relevant restaurants, dishes, items, and content to customers
Own applied machine learning problems end to end, from problem framing and data analysis to model development, offline evaluation, experimentation, and production monitoring
Develop methods that balance relevance with product and customer needs, including diversity, availability, business constraints, and changing user intent
Collaborate with Software Engineers, ML Engineers, Product Managers, and Analysts to turn scientific insights into reliable customer-facing products
Evaluate and apply state-of-the-art machine learning methods when they meaningfully improve recommendation quality, robustness, or efficiency
Contribute to applied-science practices through technical reviews, knowledge sharing, and thoughtful experimentation
требования
Substantial hands-on experience applying machine learning to real-world problems and bringing models from development into production; relevant applied research experience with a PhD is also welcome
Experience with recommendation systems, ranking, retrieval, personalisation, or closely related machine learning problems
Ability to independently turn an ambiguous customer or product problem into a well-scoped machine learning approach, make sound trade-offs, and drive it to a measurable outcome
Proficiency in Python and experience with modern machine learning frameworks and large-scale data processing
Understanding of rigorous machine learning evaluation, including offline metrics, experiment design, and interpreting online results
Ability to communicate complex technical ideas clearly and work effectively with cross-functional partners
.svg)
