Full Job Description
Credit and risk are at the heart of our business. We9re looking for a Manager of Data Scientist, a senior individual contributor who will own the full lifecycle of the ML models that power our credit risk decisions. Reporting directly to the VP of Risk, you9ll design, build, deploy, and monitor the models that determine how we approve customers, set credit amounts, predict defaults, and forecast losses. You9ll work closely with cross-functional partners across risk, engineering, finance, and sales, and you9ll have the rare opportunity to shape both the technical infrastructure and the business strategy behind it.
What You9ll Do
3 Feature Engineering: Go beyond surface-level signals - you9ll mine internal and external datasets to engineer high-signal features (DTI, PTI, payment behavior, account balance patterns) that directly improve the predictive power of production credit models. Your work here changes approval outcomes for real customers.
3 Model Development: Own the end-to-end development of strategic credit risk models, not just as technical exercises, but as tools that shape how Kafene decides who to approve, for how much, and at what risk. Approval amount sensitivity models, credit line optimization, and loss forecasting are yours to design and improve.
3 Data Preparation: Source, clean, and transform messy real-world financial data into modeling-ready datasets. You set the standard for data integrity here; this isn9t a role where someone else handles the data pipeline for you.
3 Vendor Evaluation: Evaluate third-party data vendors and scoring products on their merits, not just their sales pitch. You9ll lead cost-benefit analyses and make the call on what gets integrated into our models.
3 Research & Innovation: You9ll have the latitude to bring in new techniques from the ML literature and apply them to real credit risk problems, not POCs that die in a notebook, but improvements that ship to production.
3 Model Implementation & Validation: Partner with engineering to get your models into production accurately and efficiently. You9ll define what 4good4 looks like for model validation and push for faster, more repeatable deployment.
3 Monitoring & Maintenance: Stay close to how your models behave in the wild. You9ll lead recalibration and redevelopment when performance drifts, because you care about outcomes, not just the initial build.
3 Compliance & Governance: Navigate model risk governance, regulatory requirements, and data vendor usage policies with confidence. You9ve done this before and know that good documentation isn9t bureaucracy - it9s what keeps great models in production.
3 Cross-Functional Partnership: Translate business questions from risk, finance, and sales into modeling problems - and explain your solutions back in plain language. The best ideas here come from people who can move between the data and the boardroom.
What You9ll Bring
3 Education: Master9s or PhD in a quantitative discipline: Statistics, Mathematics, Data Science, Econometrics, or a related field. This is a modeling-first role; a software engineering background alone won9t be the right fit.
3 Experience: 5+ years working as a Data Scientist or ML Engineer with a specific focus on predictive modeling, ideally in credit risk, fraud detection, or financial analytics. Experience deploying models that affect real credit or lending decisions is what we9re looking for.
3 Technical Skills:
3 Advanced Python for statistical modeling and ML - not primarily for application development or infrastructure engineering
3 Strong SQL for data extraction and feature construction
3 Deep expertise in ML algorithms purpose-built for structured/tabular data: gradient boosting, ensemble methods, regression models, decision trees, and AutoML frameworks
3 Industry Background: Prior experience in consumer lending, fintech, or financial services is highly preferred; you should already understand what DTI, PTI, and vintage analysis mean without needing context.
3 Governance: Hands-on experience with model risk governance frameworks and working alongside validation teams; you know the SR 11-7 world and aren9t intimidated by it.
3 Communication: You can explain a gradient boosting model to a risk committee and a credit policy tradeoff to an engineer. Both matter here.
Why Kafene
3 Your models will directly determine credit outcomes for hundreds of thousands of real customers - this isn9t a recommendations engine or an internal tool; it9s the core of the business
3 You9ll own the full ML lifecycle, from raw data to production, without having to fight for scope or hand off the interesting parts to someone else
3 ML is treated as a strategic asset at Kafene, not a support function; you9ll present to executives and shape credit policy, not just fulfill tickets
3 Competitive compensation, remote flexibility, and a team that9s genuinely excited about the problems they9re solving, including the hard regulatory and data constraints that make credit risk interesting
Compensation and Benefits:
3 Salary: Earn a competitive salary $95,000-$140,000
3 Healthcare: We prioritize your well-being by covering 80% of medical, dental, and vision insurance costs, including coverage for your spouse, children, and other dependents.
3 Retirement Benefits: Begin planning for your future from day one with our 401k plan.
3 Paid Time Off: We understand the importance of work-life balance. That9s why we offer flexible paid time off days starting from day one of your employment.
