Manager, Machine Learning

Extend

5d•Onsite

About The Position

You will lead a team of ML data scientists on the Fraud and ML team, owning the development and quality of Extend's machine learning models across fraud detection, risk assessment, and identity resolution. You'll guide your team through the full data science lifecycle, from requirements and experimentation through model development, evaluation, and monitoring. You’ll partner closely with Product and Engineering on integrating ML models into our product and with our Fraud Intelligence team to continuously improve our fraud detection capabilities.

Requirements

6+ years of work experience building and deploying machine learning systems into production
2+ years experience mentoring and managing ML teams
Strong proficiency in Python and SQL
Strong understanding of ML fundamentals: model selection, evaluation methodology, feature engineering, and common failure modes
Hands-on experience with PyTorch, scikit-learn, and XGBoost (or similar gradient boosting frameworks)
Strong people leadership skills with the ability to develop ML talent
Excellent stakeholder management, with a track record of working cross-functionally to deliver results
Empathy and humility

Nice To Haves

Experience building fraud detection or risk assessment systems
Experience with cloud ML platforms, particularly AWS (e.g., SageMaker)
Experience with graph data and graph-based models (e.g., PyTorch Geometric)
Experience with model monitoring and observability tooling (e.g., Arize)

Responsibilities

Own the model lifecycle: requirements, experimentation, model development, evaluation, and model cards, partnering with ML engineers on deployment and production infrastructure
Translate business problems into well-framed ML solutions: defining what to model, what success looks like, and where ML adds value vs. simpler approaches
Design and maintain feature engineering pipelines for model development
Drive experiment design and statistical rigor: ensuring models are evaluated with sound methodology before and after launch
Monitor model quality in production, tracking performance over time, detecting data drift, and determining when to retrain
Cultivate a culture of learning and collaboration within and across partner teams
Perform design and code reviews to raise the technical excellence bar
Hire, mentor, and coach data scientists