Data Science Engineer, Payment Risk

Adobe Adobe · Enterprise · San Jose, CA

This role focuses on building and maintaining predictive decisioning models for payment risk and fraud detection. It involves data analysis, building data pipelines, and implementing ML models to mitigate fraudulent transactions and improve risk decisioning at checkout. The role also involves defining risk policies, collaborating with internal teams and external partners, and conducting A/B tests to optimize fraud prevention strategies.

What you'd actually do

  1. Mitigate fraudulent payment activity involving credit/debit cards, SEPA, PayPal, and other payment methods worldwide. Collaborate with Adobe’s third-party risk assessment partner and internal ML team.
  2. Establish Fraud Prevention capabilities based on predictive decisioning models that you will build and maintain for Adobe to align with industry best-practices.
  3. Implement improvements that raise the barriers against new threats by influencing the revision of business processes and decision-making capabilities, and through detailed auditing and validation of these processes following implementation of these changes.
  4. Define and manage risk control measurements, implement quantitative monitoring metrics, and align internal risk teams and external risk decisioning providers on risk control numeric goals, promote results-focused, data-driven data science practices.
  5. Experiment experimental build & hypothesis testing control groups

Skills

Required

  • SQL
  • Python
  • risk strategy
  • data analysis
  • fraud detection
  • payment risk
  • A/B testing
  • hypothesis testing

Nice to have

  • machine learning model development
  • model validation
  • model deployment
  • serving infrastructure
  • operationalizing models
  • risk/abuse operations
  • SEPA
  • PayPal

What the JD emphasized

  • SQL and Python skills — You’ll be writing queries, building data pipelines, debugging code, and driving insights from large, complex datasets. Proficiency in both SQL and Python is essential to operate effectively in this role (SQL/Python is a MUST).
  • Practical machine learning experience — A solid track record applying machine learning in production, including model development, validation, deployment, or serving infrastructure, is a _strong plus_. You should be comfortable operationalizing models, interpreting results, and partnering with ML/engineering teams to integrate models into risk decisioning workflows.

Other signals

  • fraud detection
  • risk strategy
  • machine learning models
  • data pipelines
  • A/B testing