AP Ndunge
AI & Data Consultant
I empower businesses to confidently navigate the cloud and implement secure AI and data architectures.
I offer tailored solutions to unlock the power of data and AI securely, especially in emerging markets.
Tool Stack
From cloud to code, I wield a powerhouse stack—AWS to PySpark, TensorFlow to Power BI—turning raw data into AI-driven gold. Let’s architect the future, one scalable, insight-packed pipeline at a time!
I’m a specialist data and AI consultant helping enterprise leaders architect high-trust, AI-ready data platforms that reduce costs, streamline decision-making, and enable scalable intelligence across teams.
With deep expertise in modern data stacks, medallion architecture, and AI integration, I deliver tailored, end-to-end solutions that unlock a single version of truth and facts—and real business value.
About Me
Expertise And Services
Data Strategy
Crafting data strategies aligned with your business goals, ensuring data-driven decision-making and competitive advantage. Unlock your business potential with data.
AI Implementation
Implementing AI solutions tailored to your specific needs, driving innovation, automation, and efficiency across your organization. Transform your business today.
Cloud Solutions
Providing comprehensive cloud solutions, from migration to management, enabling scalability, flexibility, and cost-effectiveness. Embrace the cloud with confidence.
Cloud Security Risk Assessment
Identifying and mitigating potential security risks in your data infrastructure, ensuring compliance and protecting your valuable assets. Secure your business data.
Projects
Stock Market Analysis
The project involved extracting stock and revenue data for Tesla and GameStop using yfinance and web scraping, respectively. Visualizations were created using libraries like Matplotlib or Seaborn to analyze trends and patterns.
Credit Card Fraud (CCF) Detection Model
Developed and evaluated credit card fraud detection models using Decision Trees and SVMs with Scikit-Learn and Snap ML, addressing class imbalance and measuring performance with ROC-AUC and hinge loss. Snap ML significantly improved training speed while maintaining comparable accuracy to Scikit-Learn.
Multiple Machine Learning Model Applications
Developed machine learning models for churn prediction, cancer cell classification, customer segmentation, and fuel consumption prediction using Logistic Regression, SVM, K-means, and Linear Regression. Evaluated models with metrics like Confusion Matrix, Log Loss, and MSE, using real-world datasets across telecom, healthcare, and automotive domains.
Analyzing wildfire activities in Australia
Analyzed historical Australian wildfire data to uncover trends, seasonal patterns, and regional variations using Pandas, Seaborn, and Folium for visualization and mapping. Key findings included peaks in fire area between 2010–2013, uneven fire brightness across regions, and correlations between fire radiative power and confidence levels.
Contact: +27 66 258 1276
© 2025 Ndunge Intelligence. All rights reserved.