Yusuf Jabes

Strategic Data Scientist with 5+ years of experience leveraging predictive modeling, SQL, Python, and machine learning to optimize research, financial and operational performance. Skilled in building automated data pipelines, MLOps frameworks, and Power BI executive dashboards. Adept at model risk governance, risk analytics, and translating quantitative findings into high-impact business solutions for enterprise partners.

Lead Data Scientist & Managing Director at Somo Properties Limited (2025-08 – Present)

• Designed and deployed enterprise reporting systems and performance monitoring frameworks to track real-time financial operations across high-volume portfolios, encompassing tenant rent collection, landlord advance disbursements, and property maintenance overhead.
• Built real-time Power BI dashboards and data models that transformed raw ledger records into executive decision systems, delivering custom views for shareholders, property managers, and directors on portfolio yield and operational health.
• Automated data validation and cleaning protocols by scripting data pipelines in Python (Pandas, NumPy) and SQL, eliminating reporting discrepancies across multi-source transaction feeds for landlord repayments, tenant histories, and vendor expenses.
• Formulated quantitative metrics and evaluation frameworks in R and Excel to model tenant payment behavior, calculate landlord loan ROI, and benchmark property managers on regional collection trends and recovery timelines.
• Applied statistical modeling and predictive algorithms using Python (Scikit-Learn) to forecast cash flow liquidity, assess property-level profit margins, and calculate tenant default probability scores.
• Engineered custom data collection solutions using SQL-backed APIs and ETL workflows, collaborating with regional managers to translate operational needs into scalable, centralized data pipelines.
• Led data process optimization initiatives, using Python scripting and MLOps practices to replace fragmented manual logs with automated pipelines tracking office overhead, loan amortization, and contractor repair costs.
• Leveraged predictive analytics, risk scoring, and Git version control to evaluate landlord advance risk, segment high-risk tenant cohorts, and inform strategic board-level decisions on capital allocation.

Data Scientist at International Livestock Research Institute (2021-01 – 2025-07)

• Digital Data Capture & Systems Design: Engineered end-to-end mobile data collection architectures using ODK and KoBoToolbox, streamlining data ingest and enforcing strict schema validation for large-scale agricultural datasets.
• Integrated Data Architecture: Built unified multi-institutional data management systems integrating data capture, cloud storage, and validation protocols across Kenya's dairy value chain.
• Automated Data Engineering & Orchestration: Developed scalable ETL/ELT pipelines using Apache Airflow, SQL, and Cloud platforms, reducing manual processing time and ensuring real-time data availability for analytics.
• National Data Harmonization & Governance: Orchestrated data standardization protocols in collaboration with national bodies (KALRO, DVC, ADC, and the Ministry of Agriculture) to establish uniform data models and data quality standards.
• Advanced Analytics & Statistical Modeling: Conducted exploratory data analysis and quantitative modeling using Python, R, and SQL to evaluate value chain economics and extract actionable insights for strategic policy-making.
• Business Intelligence & Analytics Visualization: Built executive dashboards and interactive reports using Python visualization packages and BI tools to translate complex statistical outputs for non-technical leadership and government partners.
• Value Chain & Economic Research: Applied microeconomic modeling and statistical techniques to assess smallholder farm profitability, providing data-driven evidence for policy intervention and yield optimization.
• Data Quality Assurance (DQA) & MLOps Hygiene: Established automated validation scripts and quality assurance frameworks to monitor data drift, missingness, and schema integrity across continuous data streams.
• Data-Driven Strategy & Recommendations: Formulated predictive and descriptive analytical insights that optimized operational resource allocation, program interventions, and multi-stakeholder strategy.
• Stakeholder Analytics Consulting: Led analytics review workshops with domain experts and policymakers, leveraging trend analysis to drive continuous operational improvement.

Bachelor of Computer Science in Computer Science – Presbyterian University of East Africa (2017 – 2023)

Kenya Certificate Of Secondary Education (KCSE) – Njoro Boys High School (2013 – 2016)