Senior Machine Learning Engineer

Affirm’s Machine Learning team solves problems critical to our business model - personalizing shopping experiences, detecting fraud, optimizing interest rates, and assessing creditworthiness in real time. Our innovative products necessitate the creation of novel machine learning solutions to drive both existing and new products.

Use Affirm’s proprietary and other third party data to develop machine learning models that predict the likelihood of fraud. These models will protect victims’ identities from being stolen, prevent Affirm from incurring financial loss, and increase the trust that consumers and partners have in the Affirm ecosystem
Partner with the ML platform team to build fraud specific ML infrastructure
Research ground breaking solutions and develop prototypes that drive the future of fraud decisioning at Affirm
Implement and scale data pipelines, new features, and algorithms that are essential to our production models
Collaborate with the engineering, fraud, and product teams to define requirements for new products
Develop fraud models to maximize user conversion while minimizing fraud losses and data costs

Bachelors in a technical field with 5+ years experience building and deploying ML models. Relevant PhD can count for up to 2 YOE
Proficiency in machine learning with experience in areas such as gradient boosting, online learning, and deep learning. Domain knowledge in fraud risk is a plus
Strong programming skills in Python
Experience using large scale distributed systems like Spark and Ray
Experience using machine learning frameworks such as scikit-learn, pandas, numpy, xgboost, and pytorch
Excellent written and oral communication skills and the capability to drive cross-functional requirements with product and engineering teams
The ability to present technical concepts and results in an audience-appropriate way
Persistence, patience and a strong sense of responsibility – we build the decision making that enables consumers and partners to place their trust in Affirm!