Senior Data Scientist, Credit Modeling

at Paypay

Hybrid, hybrid, senior, full time, 6+ years

posted 14 Nov 2023

Apply for this role Sign in to track your application

BigQuery
scikit-learn
Amazon Redshift
XGBoost
Amazon Kinesis
Kafka
Databricks
Spark
Python
SQL

About the role

Posted: 2023-11-14

About PayPay

PayPay is a FinTech company that has grown to over 75M (as of August 2026) users since its launch in 2018. Our team is hugely diverse with members from over 50 different countries.

OUR VISION IS UNLIMITED_

We dare to believe that we do not need a clear vision to create a future beyond our imagination. PayPay will always stay true to our roots and realize a vision (future) that no one else can imagine by constantly taking risks and challenging ourselves. With this mindset, you will be presented with new and exciting opportunities on a daily basis and have the opportunity to grow and reach new dimensions that you could never have imagined. We are looking for people who can embrace this challenge, refresh the product at breakneck speed and promote PayPay with professionalism and passion.

※ Please note that you cannot apply or be selected in parallel with PayPay Corporation, PayPay Card Corporation and PayPay Securities Corporation.

As a Senior Data Scientist in PayPay Credit Modeling, you will lead work that improves credit risk and financial decisions.

You will lead modeling initiatives from identifying opportunities and framing ambiguous business problems through developing models, designing credible evaluations, and assessing real-world impact. You will collaborate closely with product, risk, data engineering, and machine learning engineering teams to translate research into reliable, scalable solutions.

This is a role with scope to shape the technical direction of modeling and analysis, coach colleagues, and influence decisions across PayPay.

What you would do

- Own the scientific approach to modeling work, from defining the decision and expected value through baselines, validation, and a clear recommendation.

- Develop and evaluate models and features using large, varied datasets. Choose methods that fit the problem and can be explained, validated, and maintained.

- Set evaluation standards for how models will be used. Challenge assumptions, detect leakage and selection bias, account for delayed outcomes, and use appropriate baselines, time-aware tests, and calibration to judge whether results are reliable.

- Estimate likely effects on revenue, credit losses, exposure, portfolio quality, and customer outcomes. Make assumptions and uncertainty clear.

Read from job-boards.greenhouse.io, last seen 3d ago.

What they are looking for

- Bachelor’s Degree in a quantitative field such as Computer Science, Mathematics, Statistics, Operations Research, Machine Learning or equivalent. An advanced degree is a plus.

- Substantial professional experience in data science or applied machine learning, typically six years or more, with evidence of leading consequential modeling work from problem definition through production use and impact assessment.

- Core data science skills:

- Strong knowledge of statistical modeling, supervised learning, feature engineering, and model interpretation; sound judgment on when a simpler method is sufficient.

- Practical experience with validation and experimentation, including time-aware evaluation, calibration, bias or selection effects, and monitoring after deployment.

- Ability to work independently with large, imperfect datasets and distinguish a useful signal from leakage or artifacts.

- Financial and decision skills:

- Ability to connect model outputs with financial and customer outcomes, and explain the trade-offs in a decision or policy.

- Ability to form a recommendation when evidence is incomplete, state its limits, and identify what would change the conclusion.

- Core technical skills:

- Strong hands-on Python and SQL skills for model development, data analysis, and reproducible work.

- Experience using common statistical and machine learning libraries such as scikit-learn or XGBoost.

- Experience working with large-scale data platforms or distributed data processing to develop features, train models, and evaluate results.

- Experience building reproducible analysis and model workflows, collaborating with engineering teams on deployment, and evaluating models after release.

- Experience building point-in-time features and reproducible model workflows, and working with engineering colleagues to deploy, monitor, and improve models in production.

- Experience in lending, banking, payments, or another financial decision domain. Credit risk and portfolio economics experience is especially valuable.

- Experience applying models to financial decisions, credit strategy, or portfolio optimization.

- Experience with causal inference, treatment effects, decision policy evaluation, or controlled experiments.

- Hands-on experience with Databricks, Spark, BigQuery, or Amazon Redshift; experience with streaming tools such as Kafka or Amazon Kinesis is a plus.

Ways of working

- Take ownership of ambiguous questions and bring structure to the work needed to answer them.

- Challenge assumptions constructively and revise your view when the evidence changes.

- Balance speed with rigor, explainability, and the ability to maintain what you build.

- Communicate trade-offs and decisions clearly to partners with different expertise and priorities.

PayPay 5 senses

- Please refer PayPay 5 senses to learn what we value at work.

Working Conditions

Employment Status

- Full Time

Office Location

- Hybrid Workstyle ※ You will be expected to work in the office, in alignment with organizational guidelines and team objectives.

LIFE in JAPAN FACTBOOK

Work Hours

- Super Flex Time (No Core Time)

- In principle, 9:00am-5:45pm (actual working hours: 7h45m + 1h break)

Holidays

- Every Sat/Sun/National holidays (In Japan)/New Year's break/Company-designated Special days

Paid leave

- Annual leave (up to 14 days in the first year, granted proportionally according to the month of employment. Can be used from the date of hire)

- Personal leave (5 days each year, granted proportionally according to the month of employment) ＊PayPay's own special paid leave system, which can be used to attend to illnesses, injuries, hospital visits, etc., of the employee, family members, pets, etc...

About the role

What you would do

What they are looking for

What comes with it