Uber's AI Security team is building the foundation for dynamic, data-driven security systems and evolving Uber's Zero Trust Architecture (ZTA) to be more risk-adaptive across authentication and authorization
Requirements
- 3+ years experience building and deploying ML models in production, with hands-on work in feature engineering, training, and evaluation.
- Proficiency in Python and ML frameworks (PyTorch, TensorFlow, or similar).
- Strong foundation in ML algorithms: tree-based models (XGBoost, LightGBM), classical methods (logistic regression, SVMs), and exposure to neural networks (CNNs, RNNs, Transformers).
- Experience with risk, fraud, anomaly detection, or security-related ML systems.
- Familiarity with large-scale data/infra systems (Kafka, Hive, Spark, Flink, Pinot).
- Exposure to handling challenges such as imbalanced data, feedback loops, or iterative retraining.
- Ability to analyze business/security requirements and support translating them into ML use cases.
Responsibilities
- Support framing business and security problems as ML tasks.
- Build and iterate ML models that enable risk-adaptive, real-time decisions.
- Engineer features from Uber's risk systems, logs, and contextual signals.
- Deploy and maintain ML pipelines in production, ensuring reliability and scalability.
- Collaborate with senior engineers to integrate ML into Uber's authentication and authorization systems.
- Build models and features, and take them into production
- Translate business and security needs into concrete ML problems
Other
- 3+ years experience
- Strong communication skills and ability to work cross-functionally with infra, risk, and security teams
- Ability to work in San Francisco, CA or Sunnyvale, CA
- Eligibility to participate in Uber's bonus program
- Eligibility for various benefits
