ML Engineer II - Real-Time Underwriting (Remote)

Affirm; New York (NY); On-site; USD 165,000 - 225,000; Full time

Benefits offered by this job

Health coverage for you and dependents

Equity reward program

Monthly stipends for health, wellness,

Tech spending stipends

Comprehensive medical, dental, vision

Job summary

Affirm is building underwriting ML systems to judge repayment risk in real-time. You’ll take models from idea to production with strong monitoring and measurement as user behavior and macroeconomics evolve.

You’ll develop models for tabular and sequential data, build feature pipelines, and prototype ideas, then productionize with proper risk controls and monitoring. Strong Python and distributed computing skills are required.

Qualifications

• 2+ years in ML engineering or PhD in a related field.
• Strong Python and production-quality coding skills.
• Experience building and evaluating classification models.
• Familiarity with gradient-boosted trees and modern DL frameworks.
• Experience with distributed data processing (Spark preferred).
• Knowledge of ML lifecycle tooling (Kubeflow/Airflow/MLflow).
• Ability to write clear, well-tested, extensible code.

Responsibilities

• Develop and iterate underwriting prediction models for tabular and sequential data.
• Build feature pipelines and training datasets from diverse signals.
• Prototype new ideas, run offline experiments, and productionize top approaches.
• Integrate models into batch or real-time decision systems with reliability.
• Instrument, monitor model and data health; define retraining workflows.
• Collaborate with Engineering, Risk Analytics, Product and ML Platform to convey results.

Skills

Python

Production-grade code

ML model evaluation

Model deployment

Cross-functional collaboration

Communicate results clearly

Ownership & proactive growth

Education

Bachelor’s degree in a related field

PhD in a relevant field

Tools

LightGBM

XGBoost

CatBoost

PyTorch

Spark

Kubeflow

Airflow

MLflow

Job description

Affirm is building underwriting ML systems to judge repayment risk in real-time. You’ll take models from idea to production with strong monitoring and measurement as user behavior and macroeconomics evolve.

You’ll develop models for tabular and sequential data, build feature pipelines, and prototype ideas, then productionize with proper risk controls and monitoring. Strong Python and distributed computing skills are required.