Manager, Machine Learning Engineering (fraud)

Affirm Affirm · Fintech · United States · Remote · Checkout

Manager of Machine Learning Engineering for Fraud at Affirm, focusing on building and iterating on fraud detection models throughout the ML lifecycle, including advanced techniques like representation learning and transformers. The role involves setting strategy, leading a team, and cross-functional collaboration to integrate models into decisioning systems.

What you'd actually do

  1. Set the technical and modeling strategy for fraud detection, aligning team efforts with key business outcomes such as fraud loss reduction, approval rates, and customer experience
  2. Lead a team of machine learning engineers to design, build, and iterate on high-impact fraud models across the full ML lifecycle, from experimentation to production
  3. Drive the evolution of modeling approaches, including the adoption of representation learning, transformer-based methods, and other advanced techniques for modeling complex behavioral data
  4. Partner cross-functionally with Product, Fraud Analytics, Risk, and Engineering to define solutions, evaluate trade-offs, and ensure models are effectively integrated into decisioning systems
  5. Develop talent by coaching engineers, providing feedback, and fostering a high-performing team culture grounded in technical excellence and ownership

Skills

Required

  • Bachelor’s in a technical field
  • 8+ years of industry experience
  • 3+ years managing engineers
  • modern ML approaches
  • representation learning
  • deep learning
  • transformer-based models
  • gradient-boosted trees
  • leading teams delivering end-to-end ML solutions in production environments
  • experimentation
  • evaluation
  • model iteration in production
  • strong engineering fundamentals
  • scalable systems
  • data pipelines
  • cross-functional collaboration
  • operating in ambiguous, fast-evolving environments

Nice to have

  • publication track record

What the JD emphasized

  • managing engineers
  • end-to-end ML solutions in production environments
  • model iteration in production
  • advanced techniques

Other signals

  • fraud detection
  • ML models
  • production deployment
  • transformer-based methods