Staff Marketing Data Scientist, Machine Learning

SoFi SoFi · Fintech · San Francisco, CA · General Marketing

Staff Marketing Data Scientist at SoFi, focused on building and scaling machine learning models for marketing and growth across financial products. The role involves developing predictive models for acquisition, conversion, retention, and LTV using behavioral, transactional, and credit data, with a strong emphasis on production implementation, monitoring, and regulatory compliance. The role also involves building feature stores and experimentation frameworks.

What you'd actually do

  1. Design, develop, and deploy machine learning models to optimize customer acquisition, onboarding, and lifecycle engagement across financial products (e.g., loans, credit cards, money, invest, crypto and etc).
  2. Build predictive models for critical business outcomes, including Customer Lifetime Value (LTV), conversion propensity, cross-sell/upsell effectiveness, and retention across various channels (e.g., direct mail, email, in-app, Operations).
  3. Leverage structured and unstructured data (e.g., behavioral signals, transaction data, credit attributes) to drive audience segmentation and personalization at scale. Develop and maintain a robust feature store to accelerate the end-to-end model development lifecycle.
  4. Build experimentation frameworks to support A/B testing and measure the incremental business impact of the models. Work with the other team members to incorporate the models in marketing/campaign strategies.
  5. Collaborate with Fair Lending (FL) and Model Risk Management (MRM) to ensure the models and targeting strategies adhere to fair lending laws, privacy regulations, and responsible marketing practices.

Skills

Required

  • Python
  • SQL
  • scikit-learn
  • PyTorch
  • TensorFlow
  • experimentation
  • hypothesis testing

Nice to have

  • ML governance
  • model monitoring
  • compliance frameworks
  • LLMs
  • prompt engineering
  • AI application development

What the JD emphasized

  • end-to-end model development cycle
  • implement the model in production
  • monitor performance
  • regulatory standards
  • Fair Lending (FL)
  • Model Risk Management (MRM)
  • fair lending laws
  • privacy regulations
  • responsible marketing practices

Other signals

  • build and scale machine learning models
  • end-to-end model development cycle
  • implement the model in production
  • monitor performance
  • customer acquisition, conversion, retention, and customer lifetime value (LTV)