- Advanced Analytics: Exploring time series forecasting and deep learning applications
- Full-Stack Innovation: Building scalable web applications with modern frameworks
- Data Visualization: Creating interactive dashboards that tell compelling data stories
- AI Integration: Incorporating machine learning into practical business solutions
3rd Place Winner - Inflection AI Hackathon | Africa Impact Network
Empowering Kenyan farmers with AI-driven drought resilience insights to make informed crop decisions and reduce agricultural losses by up to 30%. Features Random Forest ML model with 70% accuracy, real-time predictions, and mobile-first design.
End-to-end ML system predicting bank customer churn with a modern web interface
Comprehensive churn prediction system featuring EDA, feature engineering, multiple ML algorithms (XGBoost, LightGBM, CatBoost), REST API for model serving, and a React dashboard for real-time predictions and model analysis. Target ROC-AUC > 0.85.
- Quick start:
pip install -r requirements.txt·npm install· run FastAPI + React - Features: Model comparison, feature importance, batch/single predictions, exportable results
AI-powered demo app for analyzing customer transactions and predicting product bundles
This platform leverages association rule mining to extract actionable product associations from retail data. Features include interactive charts, downloadable CSVs, top-10 bundle recommendations, and custom filters (customer ID, confidence). Fullstack implementation with Flask backend and React frontend.
- Association Rule Mining for frequent itemsets
- Downloadable results & interactive chart visualizations
- Built-in dark/light mode, filtering, and performance metrics