Senior Machine Learning Engineer (Research Scientist) - Fraud
Plaid · New York City Office
About this role
**Senior Machine Learning Engineer (Research Scientist) - Fraud** **About the Role** Join Plaid's Fraud Data team to lead applied research on next-generation fraud detection models. You'll partner with ML Engineers and Data Scientists to develop innovative solutions across relational graphs, sequential events, images, and video data using cutting-edge approaches like Graph Neural Networks and Transformer-based models. **Key Responsibilities** • Research and prototype state-of-the-art approaches in graph ML, sequential modeling, and multimodal learning • Own a research roadmap translating innovative ideas into production solutions with measurable impact • Publish and share applied research while collaborating across Data, Product, and Engineering teams • Leverage Plaid's network-level financial data to develop fraud detection solutions **Required Qualifications** • PhD in ML, AI, Computer Science, Statistics, Applied Mathematics, or related field (or equivalent research experience with publications/patents) • 2-4+ years of relevant industry or research lab experience with demonstrated research leadership • Track record of translating research into measurable product/business impact • Strong Python proficiency and ability to build high-quality research prototypes • Excellent written and verbal communication skills **Nice-to-Have** • Experience in fraud detection, security, risk, or abuse prevention • Experience with large-scale training, graph systems, and sequential modeling **About Plaid** Plaid powers financial connectivity for millions of users across 12,000+ institutions in the US, Canada, UK, and Europe. We're committed to building a diverse team and unlocking financial freedom for everyone. **Benefits & Compensation** Comprehensive benefits including medical, dental, vision, 401(k), equity, and competitive pay based on experience and location.
Listing freshness
CronJobs last confirmed this listing 1h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.