Uber Eats is looking to build the intelligence layer that powers personalization, relevance, and product understanding across the platform. This involves tackling complex and high-impact problems in personalization and catalog ML at Uber scale to redefine how millions of people discover and shop for food every day.
Requirements
- Programming language (e.g. C, C++, Java, Python, or Go)
- Large-scale training using data structures and algorithms
- Modern machine learning algorithms (e.g., tree-based techniques, supervised, deep, or probabilistic learning)
- Machine Learning Software such as Tensorflow/Pytorch, Caffe, Scikit-Learn, or Spark MLLib
- Deep Learning
- Scalable ML architecture
- Experience in applying machine learning models to solve large-scale real-world problems
Responsibilities
- We design models that learn from user preferences and contexts to surface the most relevant grocery items and restaurant dishes, driving both discovery and loyalty.
- We build models that identify what each item truly is-brand, flavor, attributes, and the customer segments it appeals to-turning raw merchant data into a clean and intuitive shopping experience.
- Our systems learn how items relate to one another-what's a substitute, what's often bought together, and what combinations optimize outcomes for both customers and merchants.
- We develop large-scale forecasting systems to predict availability and reduce substitutions-operating at a scale few companies reach.
- We use cutting-edge computer vision to digitize physical grocery stores in real time, enabling accurate inventory systems and powering operational excellence.
Other
- PhD or equivalent in Computer Science, Engineering, Mathematics or related field AND 4-years full-time Software Engineering work experience OR 10-years full-time Software Engineering work experience, WHICH INCLUDES 4-years total technical software engineering experience in one or more of the following areas:
- 4+ years of people management experience
- Personalization, user understanding and targeting
- Optimization (RL/Bayes/Bandits)
- Causal inference
