Senior Data Scientist

Job Description Role: Senior Data Scientist\nReports to: Chief Data & Analytics Officer (CDAO)\n\nAbout Specialty Capital\nSpecialty Capital is a data-driven alternative finance company delivering fast, flexible capital to small and mid-sized businesses. We operate at the intersection of underwriting, risk, and technology using data and machine learning to make smarter, faster funding decisions while responsibly managing credit and fraud risk.\n\nAs the company scales, we are investing heavily in building a best-in-class internal data science and AI capability to power our next phase of growth.\n\nThe Mission\nBuild Specialty Capital’s internal machine learning and risk modeling capability from the ground up. This role owns the design, development, and productionization of ML/AI models that directly influence underwriting decisions, fraud detection, default risk, and lead conversion.\nYou will act as the technical authority for applied ML across credit risk and fraud , partnering closely with leadership to translate business risk into scalable, production-grade systems with measurable financial impact.\n\nWhat You’ll Do\nModeling & Analytics (Credit + Fraud)\nDesign, develop, and validate predictive models across:\nCredit risk and default probability\nFraud detection and early-warning signals (e.g., synthetic identities, misrepresentation, repeat offenders, anomalous behavior)\nFunding capacity and deal sizing\nLead scoring and submission-to-funding optimization\nApply statistical, machine learning, and ensemble techniques (e.g., logistic regression, gradient boosting, tree-based models) with a strong focus on precision/recall tradeoffs, interpretability, and real-world cost of errors .\nDevelop approaches that balance fraud prevention, approval rates, and customer experience.\nEnd-to-End Model Ownership\nOwn the full ML lifecycle, including:\nData exploration, profiling, and quality assessment\nFeature engineering across behavioral, transactional, temporal, and alternative data\nModel training, validation, stress testing, and bias analysis\nDeployment, monitoring, and ongoing recalibration\nDefine and track KPIs across risk domains (e.g., default rate, fraud loss rate, false positives, approval rate).\nOperate with high autonomy, owning outcomes rather than executing predefined tasks.\nRisk & Data Innovation\nPartner directly with the CDAO to:\nIntegrate alternative, behavioral, and third-party data sources for both credit and fraud use cases\nExperiment with novel algorithms, hybrid rules + ML approaches, and real-time scoring frameworks\nContinuously adapt models to emerging fraud patterns while maintaining stable portfolio performance.\nInfrastructure & MLOps\nHelp design and implement the ML production stack, including:\nCloud-based deployment (AWS or Azure)\nReal-time and batch scoring pipelines\nModel versioning, monitoring, drift detection, and retraining\nContainerization and API-based model serving (Docker, REST)\nEstablish best practices for model governance, reproducibility, and risk controls.\nLeadership & Influence\nServe as the technical lead for applied ML across credit and fraud risk.\nPartner closely with underwriting, operations, and leadership teams to operationalize model outputs.\nMentor junior data scientists and analysts as the team grows.\nShape the company’s long-term ML and risk roadmap.\n\nWho You Are – Must Haves\nExperienced Risk Modeler: 3+ years in data science with hands-on experience in fintech, lending, credit risk, fraud, or MCA environments. You understand default rates, fraud loss, false positives, and submission-to-funding funnels.\nBuilder Mentality: Comfortable acting as the primary architect and implementer in a greenfield or lightly structured environment.\nStrong Technical Foundation:\nAdvanced proficiency in Python (pandas, NumPy, scikit-learn, XGBoost/LightGBM)\nStrong SQL for analytical and production workflows\nExperience with Git and collaborative development practices\nBusiness-Aware: You design models with a clear understanding of underwriting economics, fraud tradeoffs, operational constraints, and downstream financial impact.\n\nNice to Have – AI, Fraud & Advanced Tooling\nFraud-Specific Experience:\nExposure to fraud typologies, anomaly detection, network/graph-based features, or velocity rules\nExperience combining rules-based systems with ML models\nGenerative AI & LLMs:\nBuilding internal AI tools using APIs (OpenAI, Anthropic) or frameworks like LangChain\nUse cases such as merchant risk summaries, fraud review support, or underwriting policy interpretation\nExplainability & Governance:\nExperience with SHAP or similar explainability techniques\nFamiliarity with audit-ready or compliance-aware modeling in financial services\n\nWhy This Role Matters\nDirect ownership of models that control credit risk, fraud losses, and revenue growth\nHigh visibility and close partnership with executive leadership\nOpportunity to define Specialty Capital’s long-term ML and risk foundation