Role Overview
Supports development and implementation of machine learning models that detect and prevent fraud across Crypto and Debit Card portfolios. Works with senior ML engineers, data scientists, and fraud strategy teams to prepare data, build features, train and evaluate models, and help deploy and monitor them in production on real-time, high-impact systems.
Responsibilities
- Assist in developing and optimizing ML models for fraud detection across debit card transactions and crypto buy/sell/transfer activity.
- Preprocess, clean, and analyze large transactional, behavioral, and device datasets to ensure data quality.
- Build and test features such as transaction velocity and behavioral signals under senior engineer guidance.
- Train, evaluate, and compare models using fraud metrics including precision, recall, and false positive rates.
- Support deployment and maintenance of ML models in production, including pipeline updates and testing.
- Monitor deployed models for performance and drift, and flag issues to the team.
- Collaborate with data scientists, engineers, and fraud operations partners to understand fraud patterns and business needs.
- Write clean, well-tested code and contribute to model documentation.
Requirements
- 1+ years relevant experience (internships, co-ops, or academic projects count) and a Bachelor’s degree, or equivalent combination of education and experience.
- Familiarity with ML frameworks like TensorFlow, PyTorch, or scikit-learn.
- Proficiency in Python and SQL.
- Foundational understanding of machine learning concepts, statistics, and model evaluation.
- Strong analytical and problem-solving skills.