Fraud Model Developer

The role:

We are looking for a Data Scientist and/or Machine Learning model developer to join our Fraud Model Development Team. This team member’s responsibilities include model development and performance monitoring supporting data-driven decision-making in partnership with our Fraud and Risk management teams. The Data Scientist will play a key role in developing Fraud models to reduce loss, minimize false positives and ultimately help SoFi protect our members. These models will apply to all SoFi products and services including Personal Loans, Student Loans, Credit Cards and Crypto. 

The Staff Data Scientist will contribute to the performance analysis of SoFi products using empirical measurements, develop quantitative and machine learning models to minimize Fraud losses and provide insights on the drivers for losses. She/He will also collaborate with the Business Units, Finance, Accounting, Operations and Fraud Risk groups. This position requires knowledge of data analytics and modeling using Python and machine learning/analytical packages as well as strong problem solving skills. The ideal candidate should have hands-on knowledge on common Fraud reduction methodologies and excellent knowledge of data science, statistical methodologies and machine learning models (e.g. linear regression, logistic regression, decision trees, gradient boosting, random forests, neural network, clustering analysis etc.).

By joining SoFi, you'll become part of a forward-thinking company that is transforming financial services for the better. We offer the excitement of a rapidly growing startup with the stability of an industry leading leadership team.

What you’ll do:

The Data Scientist will help SoFi develop better data driven modeling solutions by:

• Developing quantitative/machine learning models to reduce Fraud losses, and OpEx related to supporting Fraud complaints and disputes


• Aggregating and synthesizing datasets from multiple data environments


• Analyzing complex datasets to understand the performance and drivers for losses across various products


• Investigating external risk data to identify trends in the market and industry


• Conducting loss sensitivity analysis


• Automating models and analytical dashboards


• Monitoring the models’ performance and re-calibrating the models as needed


• Working with Business Units, Operations, Product, Capital Markets, Finance, Accounting and Risk partners to ensure correct loss expectations and trend of losses are communicated effectively and executed appropriately

What you’ll need:

• 5+ years of loss forecasting experience and/or a Master’s or PhD degree in Statistics, Mathematics, Economics, Engineering, Computer Science, or a quantitative field


• Proficient in Python, SQL & Tableau


• Experienced in model development and data analysis with deep knowledge of data science, statistical methodologies and machine learning models, e.g. linear regression, logistic regression, decision trees, gradient boosting, random forests, neural network, clustering analysis etc.


• Hands-on knowledge on common loss forecasting methodologies


• Someone who is highly motivated and drives change, is eager to learn and able to work collaboratively in a complex and fluid environment

Nice to have:

• Familiarity working with graph databases


• Experience with developing and productionizing models in the AWS environment a plus