Data Science (Credit Risk)
Salmon is a technology-driven financial company building a banking and lending platform across Southeast Asia, starting in the Philippines.
7M+ app downloads. 2M+ monthly active users. 7000+ partner stores. US$350M+ raised from top-tier investors.
Here, you get real ownership from day one, high standards, and direct communication. You'll work alongside a team that combines deep local knowledge and experience scaling regulated financial companies from scratch.
If you want to solve complex problems at scale and directly impact how millions of people access and manage money, come build with us.
Southeast Asia's fintech moment starts here.
About the role
You'll own credit scoring across the full customer lifecycle — from application decisions to behavioral models that drive portfolio profitability
What you'll do
Build and improve Application and Behavioral Scorecards that drive underwriting and collections decisions across live credit products
Work within a regulated environment — prepare model documentation, and defend modeling decisions to internal governance and external reviewers
Collaborate with product, analytics, and engineering to translate model outputs into strategy changes and implementation specs
What you'll own
Full modeling cycle: data exploration, feature design, model build, validation, deployment specs, and post-production monitoring
Scorecard quality — tracked through AUC, KS, bad rate, stability index, and assessed against real portfolio outcomes via NPV and backtesting
Model documentation and governance artifacts ready for regulatory review
Credit strategy input — your models influence approval, pricing, and collection decisions, not just inform them
What makes you a strong fit
3+ years of hands-on experience building credit risk models — Application Scorecards and/or Behavioral Scorecards specifically (ECL/IFRS9-only experience does not substitute)
Participated in model audits, prepared documentation, or supported regulatory review
Strong Python (pandas, scikit-learn) and SQL — you build models yourself, not just review others' work
Experience covering the full lifecycle: data, feature engineering, validation, production monitoring
Familiarity with Databricks or similar environments is a plus