Project 01

Fraud Detection System for Major Bank

Developed a real-time machine learning fraud detection system for a leading South African bank, reducing fraudulent transactions by 78% within the first quarter of deployment. The system processes over 1 million transactions daily, identifying suspicious patterns and flagging high-risk activities instantly.

Technologies: Python, TensorFlow, Apache Kafka, PostgreSQL, AWS SageMaker

Project 02

Retail Customer Personalization Engine

Built an AI-powered personalization engine for a national retail chain, delivering tailored product recommendations and targeted promotions. The solution increased average order value by 45% and improved customer retention rates by 32% through data-driven segmentation and predictive modeling.

Technologies: Python, Scikit-learn, Apache Spark, MongoDB, Google Cloud AI

Project 03

Healthcare Data Lake & Analytics Platform

Designed and implemented a centralized data lake for a major healthcare provider, integrating data from 12 hospitals. Enabled real-time patient monitoring, predictive analytics for readmission risks, and streamlined regulatory reporting. Reduced patient readmission rates by 25%.

Technologies: AWS Lake Formation, Apache Airflow, Tableau, Python, Redshift

Project 04

Real-Time Data Pipeline for Telecom Operator

Architected a real-time data processing pipeline for a leading telecom operator handling over 500 million events daily. The solution enables real-time network monitoring, customer churn prediction, and dynamic pricing optimization. Achieved 99.5% pipeline uptime with sub-second latency.

Technologies: Apache Kafka, Flink, Cassandra, Docker, Kubernetes, Azure

Project 05

BI Dashboard for Government Agency

Developed a comprehensive business intelligence dashboard for a government agency, consolidating data from 20+ departments into a unified view. Enabled real-time budget tracking, resource allocation optimization, and data-driven policy decisions. Improved reporting efficiency by 60% and reduced manual data processing by 200+ hours monthly.

Technologies: Power BI, Azure Synapse, SQL Server, Data Factory, DAX

Project 06

Supply Chain Analytics for Logistics

Implemented a comprehensive supply chain analytics solution for a national logistics company, optimizing route planning, warehouse operations, and demand forecasting. The system reduced logistics costs by 30%, improved on-time deliveries by 22%, and decreased fuel consumption through intelligent route optimization.

Technologies: Python, Apache Spark, Tableau, PostgreSQL, Google Maps API, AWS

Project 07

NLP Chatbot for Insurance Claims

Developed an intelligent NLP-powered chatbot for a major insurance company, automating the claims triage and processing workflow. The chatbot handles 60% of initial claims inquiries autonomously, reducing average claims processing time from 48 hours to just 4 hours and achieving 92% customer satisfaction.

Technologies: Python, BERT, Rasa, AWS Lex, Lambda, DynamoDB, React

Project 08

Cloud Migration for Financial Services

Led the end-to-end cloud migration of on-premise data infrastructure for a financial services firm, transitioning 200+ databases and 50+ applications to AWS. Achieved 40% reduction in infrastructure costs, improved disaster recovery with 99.99% uptime, and ensured full regulatory compliance for POPIA and GDPR.

Technologies: AWS, Terraform, Docker, Kubernetes, CI/CD, Python, CloudFormation

Project 09

Predictive Maintenance for Manufacturing

Built an IoT-based predictive maintenance system for a manufacturing plant with 500+ industrial machines. Using sensor data and machine learning, the system predicts equipment failures 72 hours in advance, reducing unplanned downtime by 55% and extending equipment lifespan by 35%. Saved the client over R5 million annually in maintenance costs.

Technologies: Python, TensorFlow, MQTT, InfluxDB, Grafana, Azure IoT Hub

Project 10

Customer Churn Prediction for Insurance

Developed a customer churn prediction ML model for a insurance company, analyzing policyholder behaviour, claims history, and engagement patterns to identify at-risk customers. The solution enabled proactive retention campaigns, reducing customer churn by 35% within six months and increasing CLV by R12 million annually through targeted interventions and personalized offers.

Technologies: Python, XGBoost, Apache Spark, Snowflake, Tableau, AWS SageMaker