About the Role
Umba is hiring a Data Scientist to own two critical models that decide the company's profitable growth: customer acquisition and underwriting. On the credit side, you will build and improve scoring systems using bank statement data, payments history, CRB data, and behavioral signals. On the growth side, you will optimize marketing spend across channels like ad targeting, funnel conversion, and channel attribution.
This is an AI-native role where the team leverages tools like Claude Code, Codex, and other LLM-based systems to accelerate analysis and iteration. You will focus on defining problem specs for AI agents, validating AI-generated code, and building automated feedback loops. This is a highly technical, in-office role in Nairobi.
Responsibilities
Credit & Underwriting
- Build, deploy, and improve credit scoring models using transaction and CRB data.
- Design automated underwriting flows for both digital and sales-sourced applications.
- Implement model retraining pipelines for continuous improvement.
- Monitor model performance, drift detection, and automated alerting.
- Partner with Risk and Operations on policy thresholds and human-in-the-loop processes.
Growth & Marketing Analytics
- Optimize ad targeting (audience selection, bid strategy, creative performance).
- Analyze the end-to-end acquisition funnel (impression to repayment).
- Design and run A/B tests and build experimentation infrastructure.
- Develop actionable attribution and LTV/CAC models.
Cross-cutting & AI-Native Tasks
- Write technical specs for AI-assisted workflows.
- Use AI tools to accelerate data wrangling and feature engineering while maintaining rigorous validation.
- Extend the data platform with new sources (APIs, CRB providers, payment rails).
- Ensure data integrity for all lending-related decisions.
- Present findings to non-technical stakeholders.
Skills and Qualifications
- 4+ years of hands-on data science or applied ML experience in production environments.
- Strong proficiency in Python (pandas, scikit-learn, numpy) and SQL.
- Deep experience with classifier and regression modeling (feature engineering, model selection, calibration).
- Solid applied statistics (hypothesis testing, experimental design, selection bias).
- Experience with messy real-world financial data.
- Comfort with relational databases (Postgres / MySQL).
- Excellent communication skills to bridge technical and non-technical teams.
Preferred Attributes
- Experience in credit scoring or fraud modeling in emerging markets.
- Marketing analytics or growth experimentation experience (Meta, Google ads).
- Proficiency with AI coding tools (Claude Code, GitHub Copilot, etc.).
- Familiarity with Kenyan CRB data (Metropol, TransUnion) and fintech products.
How to Apply
Interested and qualified candidates should apply through the Umba application portal at umba.applytojob.com.
