Data Scientist

SoFi SoFi · Fintech · Frisco, TX · Risk Management

Data Scientist at SoFi focused on developing and improving machine learning and statistical models for credit risk and operational areas. This role involves collaborating with various teams, leveraging data sources, and ensuring model rigor and monitoring.

What you'd actually do

  1. Develop, implement, and continuously improve machine learning and statistical models that support various credit, risk, and operational procedures including but not limited to underwriting, portfolio management, loss mitigation, and loss forecasting, etc.
  2. Present model performance and insights to Credit, Risk, and Business Unit leaders.
  3. Proactively identify opportunities to apply advanced modeling approaches to solve complex business problems.
  4. Explore and leverage in-house and external data sources to enhance model predictive power.
  5. Collaborate with the Model Risk Management team to demonstrate models are developed with high level rigor that satisfy Model Risk Management and Governance requirements.

Skills

Required

  • Master's degree in Statistics, Econometrics, Mathematics, Operations Research, Physics, Computer Science, Engineering, or quantitative field
  • 2+ years of relevant work experience in building and implementing machine learning and statistical models
  • Excellent logic reasoning and communication abilities
  • Strong skills in writing efficient SQL queries and Python code
  • Strong sensitivity to details in data
  • Strong knowledge of databases and related languages/tools such as SQL, NoSQL, Hive, etc.
  • Demonstrated sophisticated experience in building efficient and reliable pipelines that interact with large datasets stored in SageMaker and Snowflake, automating recurring processes such as data extraction and processing, feature selection, model training, model monitoring, and generating documentation templates to support reproducibility and cross-functional collaboration.
  • Excellent knowledge of machine learning and statistical modeling methods for supervised and unsupervised learning
  • Strong programming skills in Python and machine learning libraries (e.g., sklearn, lightgbm, xgboost, pytorch, tensorflow, keras, etc.)

Nice to have

  • PhD degree
  • Experience in a lending organization
  • Experience with model documentation and delivering effective verbal and written communication
  • Experience in working closely with Product, Engineering, and Model Risk Management teams
  • Experience with AWS or GCP

What the JD emphasized

  • building and implementing machine learning and statistical models
  • building efficient and reliable pipelines
  • model development, deployment, monitoring, and model re-calibration/re-build process

Other signals

  • Develop, implement, and continuously improve machine learning and statistical models
  • Apply advanced modeling approaches to solve complex business problems
  • Collaborate with the Product and Engineering teams to improve the model development, deployment, monitoring, and model re-calibration/re-build process.