Amazon logoAmazon Externship via Extern

Building practical experience in operations and people analytics

I am gaining hands-on experience through an Amazon Operational Strategy & People Analytics externship while I prepare to start my career.

View my workMessage Me

Work samples

Work samples include deliverables and analysis produced during the Amazon Operational Strategy & People Analytics externship and other early-career projects.

  • KENYA CHILD AND MATERNAL HEALTH SUBNATIONAL ANALYSIS
    KENYA CHILD AND MATERNAL HEALTH SUBNATIONAL ANALYSIS

    Analyzed KDHS 2022 data across 47 counties, 32,156 women, and 19,530 children to map subnational maternal and child health disparities Built a composite need score combining stunting, wasting, vaccination, and wealth indicators into one replicable metric for prioritizing intervention Applied OLS regression identifying facility delivery access as a significant predictor of stunting (p=0.018), redirecting focus toward transport barriers Traced how wealth quintile shapes antenatal care uptake and delivery outcomes, revealing poverty as a structural filter on access Ranked all 47 counties by vulnerability, identifying Mandera, Wajir, and West Pokot as critical intervention zones Tools: Python, World Bank data, KDHS 2022, Global Data Lab

    PythonWorld Bank dataKDHS 2022Global Data Lab
  • INTERDISCIPLINARY ADVANCED DATA SCIENCE AND AI FOR SOCIO-ECONOMIC RESILIENCE IN KENYA
    INTERDISCIPLINARY ADVANCED DATA SCIENCE AND AI FOR SOCIO-ECONOMIC RESILIENCE IN KENYA

    Integrated the Kenya Social Accounting Matrix (2021), World Bank macroeconomic indicators (2010–2024), and UN SDG data to build a comprehensive model evaluating Kenya's economic resilience across sectors Applied ensemble ML models (Gradient Boosting, Random Forest, LSTM) and econometric techniques (SARIMAX, Lasso/Ridge regression) to produce GDP, unemployment, and poverty forecasts to 2030 Conducted structural multiplier analysis identifying ICT (1.18x) and Manufacturing (1.10x) as the highest-impact sectors for inclusive economic growth Used SHAP interpretability analysis to identify exports, FDI, and internet penetration as the primary drivers of Kenya's economic outcomes — translating black-box ML outputs into policy-readable insights Simulated targeted investment scenarios projecting 62,872 jobs created and 3.97% GDP growth, producing evidence-based policy recommendations for development planners Tools: Python (Pandas, NumPy, Scikit-learn, TensorFlow), SQL, Google BigQuery, Tableau

    PythonPandasNumPyScikit-learnTensorFlowSQLGoogle BigQueryTableau

About me

I am gaining hands-on experience through an Amazon Operational Strategy & People Analytics externship while I prepare to start my career.

I am Seana Zeyhle. I have some work experience and I'm currently taking part in the Amazon Operational Strategy & People Analytics externship as I build toward my career.

Externships

Amazon Operational Strategy & People Analytics Externship

Amazon

Experience

Data Systems & Intelligence Consultant (MIS & Web Development)

Kenya Water Security and Climate Resilience Project (World Bank & Government of Kenya) | Nairobi · APRIL 2025 - PRESENT

Machine Learning Intern

Future Interns · FEBRUARY 2026

Web Developer & IT Administrator — Data Analysis & Business Intelligence

EMA Advocates Law Firm | Nairobi, Kenya · MAY 2024 - DECEMBER 2025

Digital Marketing Executive — Campaign Analytics & Performance Data

Preeminent Technology | Nairobi, Kenya · JANUARY 2024 - APRIL 2024

Data and Administrative Intern

International Marketplace Consultants | Nairobi, Kenya · JUNE 2019 - AUGUST 2019

Education

University of North Texas (UNT)

Masters of Science Interdisciplinary Studies Advanced Data Analytics and AI - · Class of 2025

Glasgow Caledonian University (GCU)

Bachelor of Arts - Social Sciences - · Class of 2022

African Leadership University (ALU)

Leadership Program - · Class of 2019

Strathmore University

Diploma in Business and Information Technology - · Class of 2019

Skills

Operational strategyPeople analyticsData analysisProject-based externship experience
Back to work

KENYA CHILD AND MATERNAL HEALTH SUBNATIONAL ANALYSIS

Analyzed KDHS 2022 data across 47 counties, 32,156 women, and 19,530 children to map subnational maternal and child health disparities Built a composite need score combining stunting, wasting, vaccination, and wealth indicators into one replicable metric for prioritizing intervention Applied OLS regression identifying facility delivery access as a significant predictor of stunting (p=0.018), redirecting focus toward transport barriers Traced how wealth quintile shapes antenatal care uptake and delivery outcomes, revealing poverty as a structural filter on access Ranked all 47 counties by vulnerability, identifying Mandera, Wajir, and West Pokot as critical intervention zones Tools: Python, World Bank data, KDHS 2022, Global Data Lab

PythonWorld Bank dataKDHS 2022Global Data Lab
View all work
Back to work

INTERDISCIPLINARY ADVANCED DATA SCIENCE AND AI FOR SOCIO-ECONOMIC RESILIENCE IN KENYA

Integrated the Kenya Social Accounting Matrix (2021), World Bank macroeconomic indicators (2010–2024), and UN SDG data to build a comprehensive model evaluating Kenya's economic resilience across sectors Applied ensemble ML models (Gradient Boosting, Random Forest, LSTM) and econometric techniques (SARIMAX, Lasso/Ridge regression) to produce GDP, unemployment, and poverty forecasts to 2030 Conducted structural multiplier analysis identifying ICT (1.18x) and Manufacturing (1.10x) as the highest-impact sectors for inclusive economic growth Used SHAP interpretability analysis to identify exports, FDI, and internet penetration as the primary drivers of Kenya's economic outcomes — translating black-box ML outputs into policy-readable insights Simulated targeted investment scenarios projecting 62,872 jobs created and 3.97% GDP growth, producing evidence-based policy recommendations for development planners Tools: Python (Pandas, NumPy, Scikit-learn, TensorFlow), SQL, Google BigQuery, Tableau

PythonPandasNumPyScikit-learnTensorFlowSQLGoogle BigQueryTableau
View all work
Back to work

E-GOVERNANCE AND DIGITAL TRANSFORMATION: A CASE STUDY OF E-CITIZEN PLATFORM IN KENYA

Conducted a mixed-methods research study on Kenya's eCitizen e-governance platform, examining how Silicon Savannah's ICT infrastructure drives digital transformation in public service delivery Assessed platform effectiveness in modernising citizen engagement and service accessibility across urban and peri- urban populations using IBM SPSS for quantitative validation Generated data-driven insights on digital innovation adoption patterns across public sector ecosystems, contributing to understanding of Industry 4.0 applications in African governance contexts Demonstrated ability to translate complex technical and policy concepts into measurable, evidence-based research outcomes for academic and policy audiences Tools: IBM SPSS

IBM SPSS
View all work
Back to work

BANK FRAUD DETECTION SYSTEM USING MACHINE LEARNING

Built a production-ready fraud detection system across 2,512 bank transactions (9.28% fraud rate), comparing Gradient Boosting, Random Forest, and Logistic Regression models Achieved 99.80% AUC-ROC score, 100% fraud detection rate (70/70 transactions caught), and 97.22% F1-score with only 5.4% false positive rate Applied feature importance analysis identifying transaction amount, account balance, and login attempts as primary fraud indicators — model projected to prevent $195K–$260K in annual losses Tools: Python (Scikit-learn, Pandas), Statistical Analysis, Data Visualisation

PythonScikit-learnPandas
View all work
Back to work

KENYA SOCIAL MEDIA ROI AND PLATFORM PERFORMANCE ANALYSIS

Conducted a comprehensive data-driven analysis of social media platform performance across Kenya, examining ROI, engagement rates, audience churn, and urban-rural behavioural differences Consolidated platform-level, regional, and county-level metrics into a structured analytics portfolio, building a reusable BI reporting framework for marketing and business intelligence teams Applied statistical analysis to identify Instagram as the highest-performing platform by profitability and engagement, and TikTok as the dominant channel in rural Kenya — producing region-specific strategic recommendations Revealed Facebook as the lowest ROI platform with highest urban churn, providing evidence-based guidance for platform budget reallocation Tools: Python, SQL, Tableau, Power BI, Google Analytics

PythonSQLTableauPower BIGoogle Analytics
View all work