Senior Manager, Machine Learning Engineering

About the Role

As part of the ML Fraud team, the successful candidate will be at the forefront of Affirm's efforts to balance fraud loss, customer experience, and conversion. This is a remote position, offering the flexibility to work from anywhere in the US, and is an excellent opportunity for someone looking to make a significant impact in the digital payments industry.

The team's work is centered around building and improving machine learning systems. This includes developing new fraud prediction models, scaling feature pipelines, and ensuring the health and reliability of these models in production. Collaboration with data and platform teams is essential, as is the ability to prototype new modeling ideas, run experiments, and drive the implementation of the best approaches.

What You Will Do

• Lead the development of new fraud prediction models using a mix of approaches for tabular, graph, and behavioral data.
• Build and scale feature pipelines and training datasets from proprietary and third-party signals.
• Prototype new modeling ideas, run offline experiments, and drive the implementation of the best approaches into production.
• Productionize models by integrating them into decision systems and improving their reliability, latency, and operational robustness.
• Instrument and monitor model and data health, defining retraining and backtesting workflows as fraud patterns evolve.
• Identify and implement foundational improvements to the team's model-building processes.
• Collaborate across engineering, fraud analytics, product, and machine learning platform teams to define requirements, evaluate tradeoffs, and communicate results.
• Develop and maintain strong relationships with stakeholders to ensure the successful deployment of models.
• Stay updated with the latest advancements in machine learning and fraud detection, applying this knowledge to continuously improve Affirm's systems.

What We Are Looking For

• 6+ years of experience in researching, training, tuning, and launching machine learning models at scale. A relevant PhD can count for up to 2 years of experience.
• A track record of delivering high-impact machine learning models in low-latency live settings.
• Strong Python skills and experience writing production-quality code.
• Experience building and evaluating models for tabular classification problems, preferably with gradient-boosted decision trees.
• Experience with a deep learning framework, with PyTorch being preferred.
• Experience working with distributed data processing or parallel compute frameworks, with Spark being preferred.
• Experience with ML lifecycle tooling for training orchestration, experimentation, and model monitoring.
• Proficiency in using AI-powered developer tools to accelerate iteration, debugging, and code quality.
• Mastery in taking a simple problem into a solution that interacts with multiple software components.
• Comfort navigating a large code base, debugging others' code, and providing feedback through code reviews.
• A strong record of taking ownership of growth, seeking feedback, and demonstrating effective verbal and written communication skills.

Nice to Have

• Experience with Kubeflow, Airflow, or MLflow for model training, experimentation, and monitoring.
• Knowledge of containerization using Docker and orchestration using Kubernetes.
• Familiarity with cloud platforms such as AWS or GCP.
• Experience working in an agile development environment.
• Certification in machine learning or a related field.

Benefits and Perks

• Competitive salary and equity package.
• Comprehensive health, dental, and vision insurance.
• Generous parental leave policy.
• Flexible PTO and remote work arrangement.
• Access to professional development opportunities and training.
• A stipend for home office setup and internet reimbursement.
• Participation in Affirm's 401(k) plan.
• Opportunities for career growth and advancement within the company.

• Ensure your resume highlights specific examples of machine learning models you've developed and deployed, especially those related to fraud detection.
• Practice explaining complex technical concepts in simple terms to demonstrate your ability to communicate with both technical and non-technical stakeholders.
• Be prepared to discuss your experience with distributed data processing frameworks and how you've used them to improve model performance.
• Showcase projects or certifications that demonstrate your proficiency in deep learning frameworks like PyTorch.
• Emphasize any experience you have with AI-powered developer tools and how they've improved your workflow.
• Prepare to back your claims with data; for instance, if you mention improving model performance, be ready to provide metrics that demonstrate this improvement.