Hello! I am Tabitha Chelagat, a data analyst passionate about turning raw data into strategic assets. Over 3 years of experience in Banking, Insurance, and development sectors, I build automated ETL pipelines, predictive risk models, and real-time dashboards to reduce friction and boost profitability. I excel at identifying bottlenecks, engineering scalable solutions, and enabling high-fidelity decision making. Colleagues would describe me as a collaborative problem solver and clear communicator who translates complex findings into actionable insights and automated alerts for executives. I am excited by data-driven storytelling and continuous improvement through automation.…

Tabitha Chelagat

Hello! I am Tabitha Chelagat, a data analyst passionate about turning raw data into strategic assets. Over 3 years of experience in Banking, Insurance, and development sectors, I build automated ETL pipelines, predictive risk models, and real-time dashboards to reduce friction and boost profitability. I excel at identifying bottlenecks, engineering scalable solutions, and enabling high-fidelity decision making. Colleagues would describe me as a collaborative problem solver and clear communicator who translates complex findings into actionable insights and automated alerts for executives. I am excited by data-driven storytelling and continuous improvement through automation.…

Available to hire

Hello! I am Tabitha Chelagat, a data analyst passionate about turning raw data into strategic assets. Over 3 years of experience in Banking, Insurance, and development sectors, I build automated ETL pipelines, predictive risk models, and real-time dashboards to reduce friction and boost profitability. I excel at identifying bottlenecks, engineering scalable solutions, and enabling high-fidelity decision making.

Colleagues would describe me as a collaborative problem solver and clear communicator who translates complex findings into actionable insights and automated alerts for executives. I am excited by data-driven storytelling and continuous improvement through automation.

See more

Work Experience

Data Analyst at PULA
May 1, 2024 - August 14, 2026
Designed and implemented an automated end-to-end data pipeline ingesting CRM API data and accounting files from multiple sources into a SQL analytics database. Implemented scheduled ELT workflows using Python, SQL stored procedures, and job schedulers with downstream reporting in Power BI Service including KPI alerts and automated reporting. Developed and automated dashboards with real-time data refresh and KPI alerts for enrollment, retention, and engagement, accelerating executive response time by 50%. Collaborated with the Field Operations team to streamline data collection workflows, facilitating the registration and insurance of 1M+ smallholder farms in Uganda. Conducted risk and profitability analysis to identify high-risk policyholder segments, directly preventing potential revenue loss by flagging unprofitable product lines.
Data Analyst at NCBA Group
January 1, 2023 - April 1, 2024
Built predictive frameworks to identify potential loan defaulters, enabling proactive mitigation of over 1.4M in potential portfolio losses. Designed banking dashboards visualizing customer lifetime value, transaction patterns, and liquidity concentration to support data-driven VIP retention strategies. Developed an anomaly detection script to flag suspicious transaction patterns and unusual loan-to-income ratios, strengthening risk monitoring and fraud prevention protocols.
Monitoring & Evaluation Supervisor at World Health Organization
August 9, 2022 - December 16, 2022
Participated in auditing regional health data collection processes, uncovering critical bottlenecks in ODK-based data capture. Trained a team of 6 data collectors using ODK, ensuring high-quality, complete data capture across rural and urban locations. Coordinated with health facilities, government officials, and community leaders to streamline field operations.

Education

BSc in Control and Instrumentation Engineering at Jomo Kenyatta University of Agriculture and Technology
September 1, 2018 - September 1, 2022

Qualifications

Google Data Analytics – Coursera
February 1, 2022 - December 1, 2022
IBM Machine Learning Professional – Coursera
March 1, 2025 - December 28, 2025

Industry Experience

Financial Services, Healthcare, Non-Profit Organization
    Banking Portfolio Risk & Asset Monitoring Dashboard
    📌 Project Overview This project establishes a real-time operational monitoring framework for a retail bank managing 1,000 customers. It focuses on institutional liquidity and credit health by tracking Assets Under Management (AUM) and loan performance against strict regulatory floors. The solution leverages the Power BI Service to automate the reporting lifecycle and trigger proactive risk alerting via Microsoft Teams. 🧠 Business Questions Answered Capital Floor Monitoring: Are we maintaining our required $75M AUM floor necessary for institutional stability? Risk Exposure Tracking: Is the current Default Rate exceeding our 7% maximum ceiling, and what is the resulting dollar value of capital at risk? Liquidity Balance: Does our total loan volume of $30.33M provide sufficient revenue to offset the current asset shortfall? Geographic & Segment Performance: Which cities and risk tiers are the primary drivers of the current 13% default rate? 🎯 Strategic Recommendations Immediate Credit Tightening: Halt new originations for high-risk segments until the Default Rate trends back toward the 7% goal. Incentivized Deposit Growth: Launch high-yield asset acquisition campaigns to bridge the $13.72M AUM gap and restore capital stability. Regional Risk Audit: Use the City Performance visuals to identify and cap lending in regional "hotspots" contributing to the 13% default rate. 🔄 BI Automation & Proactive Alerting To eliminate manual monitoring and ensure 24/7 oversight, the following automation has been deployed: Scheduled Refresh: Dataset is connected to Google Drive via the Web Connector, with an automated refresh daily at 6:00 PM. Stakeholder Subscriptions: Risk Committee receives an automated executive PDF every Monday at 9:00 PM. Real-Time Microsoft Teams Alerts: Triggered immediately when critical thresholds are reached: Alert: Default Rate drops to 7% (Achievement Notification). Alert: Active Loans drop to $15M (Critical Liquidity Floor). Alert: AUM reaches $75M (Capital Floor Achievement). 🛠️ Technical Stack Data Sourcing: Google Drive integrated via Web Connector for seamless cloud-to-cloud data transfer. ETL & Cleaning: Utilized Power Query to perform advanced data transformation, handling null values and normalizing credit tiers. Power BI: Star Schema data modeling and advanced DAX for portfolio KPIs. Power Automate: Integration for automated stakeholder alerting.…
    Insurance Claims Risk & Fraud Detection System

    Insurance Claims Risk, Customer Profitability & Fraud Detection
    📌 Project Overview
    This project demonstrates advanced SQL analytics within the insurance industry. It focuses on identifying high-value customers, assessing policy risk through loss ratio analysis, and detecting suspicious claim patterns. The project utilizes a robust database schema to bridge the gap between technical data management and actionable business intelligence.

    📊 Business Questions Answered
    Who are our most valuable policyholders? Identified top-tier customers by lifetime value (LTV) for retention programs.

    Which policies are profitable vs. unprofitable? Calculated loss ratios to find segments where payouts exceed premiums.

    Who files unusually many claims? Detected potential opportunistic fraud through frequency analysis.

    What are claim trends over time? Analyzed seasonal volatility to assist in cash reserve planning.

    How do we optimize data delivery for dashboards? Implemented materialized views and strategic indexing for scalable reporting.

    💡 Key Findings & Insights
    Revenue Leaders: Abigail Martinez ($57,223), Robert Davis ($53,735), and Isabella Lopez ($48,348) represent the highest revenue-generating segment.

    Risk Mitigation: High-risk outliers like Avery Rodriguez were identified with loss ratios exceeding 200%, signaling an immediate need for underwriting review.

    Fraud Detection: The analysis flagged 154 claims filed within the first 30 days of policy activation, a major indicator of pre-existing damage.

    Anomaly Alerts: Policy IDs 39 and 356 were flagged for filing 7 claims each, significantly above the statistical norm.

    Transaction Patterns: 38.7% of high-value customers prefer Credit Card payments, suggesting an opportunity to increase “Auto-Pay” enrollment.

    🛠️ Technical Implementation
    Data Quality Layer: Implemented rigorous checks for NULL values, duplicate records, and logical date validation to ensure financial accuracy.

    Advanced Analytics: Utilized Window Functions (RANK, LAG, PERCENT_RANK) for complex customer segmentation and trend analysis.

    Performance Optimization: Created B-Tree indexes and Materialized Views to ensure sub-second query response times for executive dashboards.

    Robust Calculations: Applied NULLIF and COALESCE to prevent division-by-zero errors and handle missing data gracefully.

    📂 Project Structure
    insurance_sql_portfolio.sql: The main script containing schema creation, data validation, and analytical queries.

    policyholders.csv, policies.csv, claims.csv, payments.csv: The raw datasets used for the analysis.

    🎯 Conclusion
    This project provides a clear roadmap for transitioning from reactive claims processing to proactive risk management. By identifying high-LTV customers and automating fraud triggers, the organization can significantly reduce operational overhead and improve the overall health of its insurance portfolio.