Senior / Principal Data Scientist – Causal Inference & Forecasting on Credit Card Spend Data

Summary

Senior / Principal Data Scientist – Causal Inference & Forecasting on Credit Card Spend Data

Two openings: Senior and Principal · Hands-on individual contributor roles · Remote (US)

WHY THIS ROLE

Our client, a consumer data and analytics company serving brands and agencies, turns large-scale credit card transaction data into insight they act on. Its quantitative data science team owns the hardest statistical questions in that data: what actually caused a change in spending, how to make an imperfect panel represent the real population, and what consumers will spend next.

These are not off-the-shelf problems. You will design custom methods, ship them to production, and defend them to sharp, technical stakeholders. If you want to stay deeply hands-on in modeling rather than move into people management, this is built for you.

WHAT YOU WILL WORK ON

• Causal measurement: Measure whether marketing campaigns actually cause additional consumer spending, using synthetic controls, uplift modeling, incrementality testing and experiment design.
• Panel weighting and bias correction: Make a transaction panel representative of the population through weighting, calibration and sample-bias correction, at a sensible cost.
• Predictive modeling and forecasting: Forecast consumer spend and model propensity, churn and behavior, including behavioral embeddings built from transaction histories.
• Privacy-preserving analytics: Work within clean rooms, differential privacy and aggregation thresholds without losing statistical rigor.
• Production ML: Take models from prototype to governed, monitored pipelines in Python, SQL and a cloud data warehouse.

WHAT WE ARE LOOKING FOR

• Hands-on experience with credit card transaction data (required). You have built models on card spend data, not only analyzed it.
• Direct, hands-on depth in at least one of three focus areas: causal inference/experimentation, sample weighting/bias correction, or predictive modeling/forecasting of consumer behavior. Adjacent work does not substitute.
• A track record of solving complex, non-trivial problems with methods you designed or substantially adapted, not just applied.
• Strong statistics and machine learning on large tabular or behavioral data.
• Production-grade Python and SQL; experience with cloud data warehouses and large-scale ML pipelines.
• Clear, concise communication: you can explain and defend a methodology to a technical audience in a few sentences.

Senior level: a proven hands-on record in one or more focus areas; you own projects end to end.

Principal level: 10+ years in ML/statistics; you also set methodology across the team and mentor senior scientists, while staying hands-on.

PROBABLY NOT THE RIGHT FIT IF

• Your credit card data experience is mainly in credit risk, fraud, AML, or loss/reserve forecasting. These are valuable skills, but the problems and tech stack differ from this team's.
• You have mainly moved into managing teams and are no longer building models yourself.

Keywords: ML Scientist, Statistical Modeling, Predictive Modeling, Paneling, Normalization, Cohorting, Causal Inference, Synthetic Control, Uplift Modeling, Transaction Data, Behavioral Modeling, Forecasting, Representation Learning, Privacy Preserving ML, Cleanrooms, Production ML Pipelines

Job Description

Senior / Principal Data Scientist – Causal Inference & Forecasting on Credit Card Spend Data

Two openings: Senior and Principal · Hands-on individual contributor roles · Remote (US)

WHY THIS ROLE

Our client, a consumer data and analytics company serving brands and agencies, turns large-scale credit card transaction data into insight they act on. Its quantitative data science team owns the hardest statistical questions in that data: what actually caused a change in spending, how to make an imperfect panel represent the real population, and what consumers will spend next.

These are not off-the-shelf problems. You will design custom methods, ship them to production, and defend them to sharp, technical stakeholders. If you want to stay deeply hands-on in modeling rather than move into people management, this is built for you.

WHAT YOU WILL WORK ON

• Causal measurement: Measure whether marketing campaigns actually cause additional consumer spending, using synthetic controls, uplift modeling, incrementality testing and experiment design.
• Panel weighting and bias correction: Make a transaction panel representative of the population through weighting, calibration and sample-bias correction, at a sensible cost.
• Predictive modeling and forecasting: Forecast consumer spend and model propensity, churn and behavior, including behavioral embeddings built from transaction histories.
• Privacy-preserving analytics: Work within clean rooms, differential privacy and aggregation thresholds without losing statistical rigor.
• Production ML: Take models from prototype to governed, monitored pipelines in Python, SQL and a cloud data warehouse.

WHAT WE ARE LOOKING FOR

• Hands-on experience with credit card transaction data (required). You have built models on card spend data, not only analyzed it.
• Direct, hands-on depth in at least one of three focus areas: causal inference/experimentation, sample weighting/bias correction, or predictive modeling/forecasting of consumer behavior. Adjacent work does not substitute.
• A track record of solving complex, non-trivial problems with methods you designed or substantially adapted, not just applied.
• Strong statistics and machine learning on large tabular or behavioral data.
• Production-grade Python and SQL; experience with cloud data warehouses and large-scale ML pipelines.
• Clear, concise communication: you can explain and defend a methodology to a technical audience in a few sentences.

Senior level: a proven hands-on record in one or more focus areas; you own projects end to end.

Principal level: 10+ years in ML/statistics; you also set methodology across the team and mentor senior scientists, while staying hands-on.

PROBABLY NOT THE RIGHT FIT IF

• Your credit card data experience is mainly in credit risk, fraud, AML, or loss/reserve forecasting. These are valuable skills, but the problems and tech stack differ from this team's.
• You have mainly moved into managing teams and are no longer building models yourself.

Keywords: ML Scientist, Statistical Modeling, Predictive Modeling, Paneling, Normalization, Cohorting, Causal Inference, Synthetic Control, Uplift Modeling, Transaction Data, Behavioral Modeling, Forecasting, Representation Learning, Privacy Preserving ML, Cleanrooms, Production ML Pipelines