The posting
Lead the development, validation, and governance of the full suite of risk models for a high-growth payment product in Mexico, directly shaping the analytical foundation that drives every business decision across the product lifecycle. If you thrive on building scorecards from the ground up — starting with external data and expert-rule frameworks when proprietary data does not yet exist, then architecting the transition to custom models as portfolio mature — this is your opportunity to establish and scale a risk modeling function from inception. As Sr. Data Scientist, you will own the end-to-end model roadmap spanning origination, behavioral, delinquency propensity, collections, recovery, analytical fraud, and reserve input models. You will define development standards, documentation practices, versioning protocols, and monitoring frameworks with the rigor required to withstand audit scrutiny and regulatory review. Your work is both tactical and strategic: you will manage complex, cross-functional model programs with negotiated and defined obligations, influencing roadmaps, business and technology decisions, and model governance across the organization. Success in this role is measured by discriminatory power (KS, Gini) and its stability over time, Population Stability Index (PSI) of variables and scores, development-to-implementation cycle time, and the absence of validation findings. In a new product, the first 12 to 18 months are typically operated with generic or bureau scorecards and expert rules; this role requires someone who knows how to launch without proprietary data and plan the deliberate transition to custom models built on proprietary performance data as vintages mature.
Key job responsibilities You will own and execute the model roadmap covering origination scorecards, behavioral scorecards, delinquency propensity models, collections and recovery models, analytical fraud models, and reserve inputs (PD, LGD, EAD). You will manage the full model lifecycle: development, validation, approval, implementation, ongoing monitoring, recalibration, and retirement. You will develop the origination scorecard, including reject inference methodology, and lead feature engineering across external data sources, and partner transactional data. You will build behavioral scorecards,payment propensity, contactability, and recovery models and produce PD, LGD, and EAD inputs for reserves calculation. You will own model validation end-to-end: conducting conceptual soundness reviews, replicating results, performing stress testing and stability analysis, and issuing formal opinions with findings, severity classifications, and remediation timelines. You will produce complete technical documentation and model files to the standard required for third-party review.
A day in the life You will join working sessions with the risk team to translate champion/challenger experiment results into model refinement priorities, and calibrate payment propensity and contactability models against real-world recovery outcomes. You will coordinate with engineering to validate that deployed model versions match approved specifications and prepare the model performance package for the credit and risk committees.
About the team We are a team at Amazon building a new business, and we sit within the Payments organization. Today the team is small by design, which means you will have an outsized role in shaping our credit risk function, processes, and culture from day one. Our closest partners span finance, technology, and commercial teams, and together we are investing in creating a world-class underwriting and credit risk capability. If you want to help define something new at Amazon and see your work directly reflected in the product, we would love to hear from you.
- 5+ years of data querying languages (e.g. SQL), scripting languages (e.g. Python) or statistical/mathematical software (e.g. R, SAS, Matlab, etc.) experience - 4+ years of data scientist experience - Experience with statistical models e.g. multinomial logistic regression
- 2+ years of data visualization using AWS QuickSight, Tableau, R Shiny, etc. experience - Experience managing data pipelines - Experience as a leader and mentor on a data science team
Our inclusive culture empowers Amazonians to deliver the best results for our customers. If you have a disability and need a workplace accommodation or adjustment during the application and hiring process, including support for the interview or onboarding process, please visit https://amazon.jobs/content/en/how-we-hire/accommodations for more information. If the country/region you’re applying in isn’t listed, please contact your Recruiting Partner.
