The ad says“Strong understanding of the strengths, limitations, and applications for both traditional and modern ML methods, including gradient-boosted trees and neural networks”
The ad says“Strong understanding of the strengths, limitations, and applications for both traditional and modern ML methods, including gradient-boosted trees and neural networks”
The ad says“Strong Python skills, SQL proficiency for working with training and evaluation data, and hands-on experience with ML frameworks such as PyTorch, scikit-learn, XGBoost, or equivalents”
The ad says“Strong Python skills, SQL proficiency for working with training and evaluation data, and hands-on experience with ML frameworks such as PyTorch, scikit-learn, XGBoost, or equivalents”
The ad says“Strong Python skills, SQL proficiency for working with training and evaluation data, and hands-on experience with ML frameworks such as PyTorch, scikit-learn, XGBoost, or equivalents”
The ad says“Strong Python skills, SQL proficiency for working with training and evaluation data, and hands-on experience with ML frameworks such as PyTorch, scikit-learn, XGBoost, or equivalents”
The ad says“Strong Python skills, SQL proficiency for working with training and evaluation data, and hands-on experience with ML frameworks such as PyTorch, scikit-learn, XGBoost, or equivalents”
Values
The ad says“Experience applying newer modeling approaches, such as learned representations, transformers, or foundation models, to improve a production ML use case”
The ad says“Strongly preferred: Fraud or risk modeling experience, including familiarity with fraud patterns, delayed feedback, and the tradeoff between fraud detection and legitimate-user friction”
The ad says“Experience using graph-based systems to extract predictive signals, uncover fraud patterns, and improve fraud model performance”
The ad says“Experience applying newer modeling approaches, such as learned representations, transformers, or foundation models, to improve a production ML use case”
As a Senior Machine Learning Engineer on Plaid’s Fraud Data team, you will develop models that improve fraud detection for our customers. You will identify predictive patterns in Plaid’s network data and lead projects from initial experiments through model deployment and ongoing improvement
Investigate fraud patterns and model errors to identify new signals, improve detection, and expand coverage across customers and use cases
Develop training datasets and predictive features, addressing challenges such as incomplete labels, class imbalance, data leakage, and changing fraud behavior
Design, train, and tune models using traditional and modern ML methods, including gradient-boosted trees and neural networks, and evaluate newer architectures against existing approaches
Design experiments to test features and models, comparing performance across time periods and customer segments using agreed detection and false-positive metrics
Build data and training pipelines that support reproducible experiments and efficient iteration on features and models
Deploy models with Engineering and ML Infrastructure partners, balancing detection quality, latency, cost, and reliability
Independently lead ML projects, agreeing on priorities and evaluation metrics with Data Science and Product and coordinating work through model release
Build hands-on machine learning expertise across the full ML lifecycle, from feature engineering and experimentation to model deployment
Take models from initial experimentation through production and evaluate their impact using real-world customer outcomes
Develop experience building and scaling reliable ML systems in production
Explore how LLMs and Generative AI can improve fraud detection, prevention, and investigation
Accelerate your career in a fast-paced environment with opportunities to take ownership, solve complex problems, and make a meaningful impact
Benefits
Vibrant offices in SF, NYC, and Raleigh-Durham—with catered meals, happy hours, and clubs to keep you connected
Competitive pay, comprehensive health benefits, and support for fertility, mental health, and parental leave
Lifestyle perks including home office stipends, daycare support, and commuting benefits like CitiBike and Lyft
Preferred experience
Strong Python skills, SQL proficiency for working with training and evaluation data, and hands-on experience with ML frameworks such as PyTorch, scikit-learn, XGBoost, or equivalents
Experience constructing training datasets and addressing label quality, data leakage, class imbalance, and generalization across time periods or populations
7+ years of professional experience in machine learning, applied science, or software engineering for ML, including hands-on model development and deployment
Experience independently leading ML projects from an open-ended problem through deployment, coordinating requirements and model releases with Data Science, Product, and Engineering
Strong ML and statistical fundamentals, including feature engineering, experiment design, model evaluation, and diagnosing why a model underperforms
Hands-on experience designing, training, tuning, and deploying models, and measuring improvements in production performance or business metrics
Strong understanding of the strengths, limitations, and applications for both traditional and modern ML methods, including gradient-boosted trees and neural networks
We recognize that strong qualifications can come from both prior work experiences and lived experiences. We encourage you to apply to a role even if your experience doesn’t fully match the job description
Strongly preferred: Fraud or risk modeling experience, including familiarity with fraud patterns, delayed feedback, and the tradeoff between fraud detection and legitimate-user friction
Experience developing models that generalize across customers with different data and behavior patterns
Experience using graph-based systems to extract predictive signals, uncover fraud patterns, and improve fraud model performance
Experience applying newer modeling approaches, such as learned representations, transformers, or foundation models, to improve a production ML use case
