Hellen Kitheka

Analytical and impact-driven professional with 5+ years’ experience in Data Scient and Programme Monitoring, Evaluation and Learning (MEL) in development finance with interest on the banking and financial services sector. Currently pursuing an MSc in Data Science, with hands-on skills in Python, SQL, NoSQL, Power BI, and R. Adept at designing data frameworks, analyzing large datasets, and transforming data into actionable insights. Demonstrated success in stakeholder engagement, Key Performance Indicator (KPI) reporting, financial forecasting, and dashboard creation—now applying these skills to predictive modeling, customer analytics, credit risk, and fraud detection in a data-driven environment.

• May - July 2016, Industrial Attachment, Kenya Medical Research Institute (KEMRI)/CPHR
• Sep - Oct 2015, Evaluation Enumerator and Data Clerk, World Vision International, Mutonguni
• May - Aug 2015, Intern, Kitui County Treasury

1. Power BI and Tableau Performance Tracking and Monitoring Dashboards

Designed and deployed live dashboards to track program performance tracking and monitoring indicators, financial disbursements, and real-time field data.

2. Telecom Customer Churn Dataset

Built a customer churn prediction model using Python, applying machine learning algorithms (Naïve Bayes, SVM, Logistic Regression, Decision Tree, Random Forest), with data preprocessing, exploratory analysis, and model evaluation to improve accuracy and support business decision-making.

3. Stock Market Prediction Using LSTM and Sentiment Analysis

Developed a hybrid deep learning model using Long Short-Term Memory (LSTM) to predict stock trends on the Nairobi Securities Exchange, integrating financial time series data with sentiment analysis from news and social media. Improved predictive accuracy over traditional models (ARIMA, GARCH) by capturing both technical patterns and investor sentiment.

4. ETL and Data Warehousing Mini-Project (SQL)

Designed and implemented an ETL pipeline using SQL to extract, transform, and load customer records into a normalized schema. Focused on banking customer segmentation and product performance analysis.

5. Transaction Anomaly Detection (PySpark, Academic Simulation)

Used PySpark to process simulated transactional data and detect anomalies via clustering. Demonstrated Spark-based processing on large-scale data as part of a cloud computing course project.

• Homa Bay International Investment Conference 2024
• Blue Economy Investment and Innovation Summit

• Advisory Diamonds Awards in Digitizing Business - PricewaterhouseCoopers
• Public Speaking Program – Dress to Impress Group

• Leading Teams Training Workshop
• Future Leaders Programme – Effective Enrollment Training
• Engagement Economics Training
• Future Leaders Programme- Effective Engagement Training

• PRINCE2 Foundation and Practitioner Certified
• Certificate in Monitoring and Evaluation
• Diploma in PowerPoint, Excel, Data Literacy, Kubicle
• Diploma in Generative Artificial Intelligence
• Statistical Package for Social Science (SPSS), KESAP College
• Proposal Writing Skills for Rural Development Training, Tropical Institute for Sustainable Development.