Data Scientist, Findigs

• 4+ years of hands-on data science or applied ML experience (fintech, proptech, or other high-stakes decisioning environments preferred)
• Strong Python skills (pandas, scikit-learn, statsmodels or equivalent); this is a coding role
• Ability to design, run, and interpret A/B tests independently
• Strong SQL skills and comfort working in a modern data stack (dbt, Snowflake, Sigma, or similar)
• Solid grounding in supervised learning fundamentals (classification, regression, tree-based methods)
• Strong written communication and the ability to explain model behavior and tradeoffs to non-technical partners (e.g., PMs, CSMs)
• Intellectual curiosity about housing and credit data in particular

Desirable

• Findigs runs an AI underwriting engine (DecisionAssist) that makes or influences thousands of rental decisions every week
• As Data Scientist at Findigs, you will strengthen our data science and applied machine learning depth: owning hands-on model development, experimentation design, and ML-adjacent analysis that directly impacts renter and property manager outcomes
• Reporting to the Lead Analytics Engineer, this is a highly technical, high-ownership role for a data scientist who wants to build and improve production models, bring statistical rigor to product decisions, and grow into broader strategic scope as the team evolves
• You will partner closely with Product and Engineering to translate real-world rental risk and behavior into models, experiments, and clear insights
• DecisionAssist model development: Own feature engineering, model iteration, and evaluation for DecisionAssist. You will work across two surfaces: (1) operational model work in the DA/CAV1 serving layer, and (2) analytics-focused modeling in Snowflake for experimentation and research, as well as partner with Product and Engineering on what signals matter and why
• Experimentation and A/B testing: Design and analyze experiments across underwriting, renter-facing, and PMC-facing product changes, and bring statistical rigor and clear recommendations
• Predictive and risk modeling: Build and maintain models used in screening logic (e.g., delinquency risk, income estimation, fraud signals)
• ML infrastructure: While you won’t own the warehouse or pipeline architecture, you should be comfortable writing clean Python, working in dbt, and operating in a modern data stack
• Research and analysis: Tackle high-impact, ad-hoc questions from Product and Customer teams; e.g., what’s driving approval-rate variance, which cohorts behave differently, and what a given signal actually predicts