Upstart is looking to ensure that the risk of machine learning models is well-understood, monitored, and mitigated to maintain credibility with external stakeholders and make better decisions.
Requirements
- Basic understanding of AI/ML methodologies such as tree-based models and neural networks
- Experience coding in R, Python, or similar languages such as Matlab
- Understanding of advanced AI/ML topics such as model monitoring, fairness, and explainability
- Advanced coding skills in R, Python, and SQL, and experience using Git
- PhD in a quantitative field of study such as statistics, econometrics, finance, mathematics; or a related discipline
- 1+ years of experience in model risk management or model governance, or related fields such as ML and Data Science, Risk, Trust and Safety, or Technical Writing
Responsibilities
- Partner with model developers in the Machine Learning team to maintain robust and up-to-date model inventory, model risk assessments, documentation, and monitoring reports
- Partner with ML teams to provide effective challenge of Upstart’s models and communicate modeling assumptions to external stakeholders
- Help answer model questions from lending partners, in collaboration with other teams
- Collaborate with central Upstart’s Legal and Compliance teams as needed to ensure our model risk management implementation meets or exceeds Upstart’s internal policies as well as regulatory requirements
- Track team goals and progress. Perform risk assessment for the model risk management program
Other
- Master’s degree in quantitative field such as finance, mathematics, economics, statistics or a related discipline
- Internship or project experience related to model risk management, model governance, machine learning, or data science
- Strong communication skills: ability to adapt technical information to varying needs and audiences, and managing trade-offs such as providing modeling detail while protecting intellectual property
- Proactive mindset with the ability to take initiative
- Interest in or knowledge of consumer lending
- Travel requirements: majority of work can be accomplished remotely, but in-person sessions may be required once or twice per quarter for 2-4 consecutive days at a time
