Staff AI Scientist

What you'll do

• Introduce novel machine learning techniques to improve net credit loss and loss forecasting accuracy.
• Design auditable and reproducible AI-first lending systems, including agentic flows for document intake and policy reasoning.
• Determine the optimal use of LLM fine-tuning versus prompting based on latency, cost, and risk standards.
• Architect multi-stage agent pipelines and reusable science infrastructure to streamline experiment execution.
• Solve complex research problems such as learning from censored repayment outcomes and causal pricing optimization.
• Transfer research outputs into production underwriting and pricing systems while establishing rigorous research standards.
• Define key success metrics and balance trade-offs between approval rates, expected losses, and portfolio health.
• Mentor senior scientists and provide technical feedback to guide them in structuring reliable agent-executable components.

What we're looking for

• BS, MS, or PhD in Computer Science, Statistics, Mathematics, Economics, or a quantitative field (or equivalent experience).
• 8+ years of experience building, deploying, and leading cross-functional production ML systems.
• Proven track record of taking complex, high-stakes ML problems from concept to production with measurable business impact.
• Experience designing scalable data and feature platforms with versioning, quality contracts, lineage, and reproducibility.
• Strong communication skills to align cross-functional stakeholders and communicate model designs to regulatory standards.
• Experience mentoring senior scientists and engineers.
• Deep domain expertise in credit risk, underwriting policy, limit assignment, risk-based pricing, or loss forecasting (preferred).
• Hands-on experience with LLMs and agentic systems, including orchestration, fine-tuning, RAG, and evaluation (preferred).