SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 135,000 - 165,000 / annual
Extend is a post-purchase commerce platform that uses AI-driven solutions to help retailers enhance customer satisfaction and drive revenue growth. The company provides automated customer service, returns/exchange management, fulfillment automation, and fraud detection. Extend works with over 1,000 merchant partners across fashion, cosmetics, furniture, jewelry, electronics, and other industries, and is backed by prominent technology investors with headquarters in San Francisco.
As a Senior Machine Learning Data Scientist on the Fraud & Machine Learning team, you will own the full lifecycle of machine learning models that detect and prevent fraud, assess risk, and unlock business value from hundreds of millions of user transactions. You'll drive requirements gathering, feature engineering, model development, evaluation, and production monitoring. Your work will directly impact Extend's ability to stop fraudulent actors while maintaining a seamless experience for legitimate customers.
Key responsibilities include: owning the complete model lifecycle from conception through deployment; translating complex fraud patterns into well-framed ML solutions; designing and maintaining feature engineering pipelines; monitoring model quality and detecting data drift in production; and partnering with Product, Engineering, Fraud Intelligence, and leadership teams to define and execute fraud prevention strategies. You'll champion a culture of experimentation and collaboration across the data science organization.
Required qualifications: Bachelor's degree or higher in a quantitative field (Mathematics, Statistics, Computer Science, Engineering, Operations Research, Physics, or related); 3+ years of hands-on experience building and deploying machine learning systems to production; strong proficiency in Python and SQL; deep understanding of ML fundamentals including model selection, evaluation methodology, feature engineering, and common failure modes; hands-on experience with PyTorch, scikit-learn, and XGBoost or similar gradient boosting frameworks; high attention to detail and intellectual curiosity; understanding of user behavior and fraud patterns; collaborative team player mindset; must be located in continental United States.
Preferred qualifications include experience building fraud detection or risk assessment systems, cloud ML platforms (especially AWS SageMaker), graph data and graph-based models (PyTorch Geometric), and model monitoring/observability tools (Arize).
