KWAZI BUTHELEZI
Midrand,Gauteng
Summary
Data scientist delivering data-driven solutions across customer personalization, fraud analytics, and generative AI. Develops and deploys AI and machine learning systems using Amazon Bedrock, Python, SQL, and big data technologies. Analyzes unstructured customer interaction data to identify fraud patterns and conversion opportunities.
Overview
1
1
Certification
4
4
years of professional experience
Work History
Data Science Specialist (CX & CVM)
Vodacom Blvd
Midrand
02.2025 - Current
- Designed and developed a Digital Marketing Agent using Google Agent Development Kit (ADK) to enable intelligent retrieval and analysis of Customer 360 data for targeted marketing strategies.
- Built SR intent model with Amazon Bedrock foundation models to classify and analyze escalated customer service requests, improving email response accuracy.
- Conducted gambling behavior analysis to identify patterns and anomalies, generating actionable insights for business decisions.
Junior Data Scientist (Fraud & CX)
Vodacom Blvd
Midrand
02.2024 - 01.2025
- Supported efforts to reduce fraud-related credit note losses, addressing one of organization’s main sources of financial fraud.
- Contributed to BTS fraud model that predicted likelihood of battery theft at sites, supporting targeted security actions.
- Managed end-to-end development, deployment, monitoring, and ongoing improvement of Compromised Customer Model (CCM).
- Built and tuned machine learning models for fraud detection, improving identification accuracy, operational efficiency, and fraud prevention capabilities.
- Designed and integrated AI solutions using Amazon Bedrock, connecting foundation models through API integrations to analyze customer interactions, improve response quality, and provide business insights.
- Analyzed API calls to evaluate functionality and performance metrics. and response behavior in depth to assess performance, latency, and integration efficiency across GenAI services.
- Developed call centre-focused Customer 360 analytics to improve FCR for DRC.
Intern Big Data
Vodacom Blvd
Noordwyk
02.2023 - 01.2024
- Supported development of big data models for J4U use case using machine learning to help optimize voice and data offer distribution for customers.
- Built and presented interactive dashboards in Microsoft Power BI to improve data visualization and support business stakeholder storytelling.
Education
B.Sc. Honours - Physics
University of the Witwatersrand
Johannesburg, South Africa12-2022
B.Sc. - Mathematics and Physics
University of the Witwatersrand
Johannesburg, South Africa12-2021
Skills
SQL proficiency
API integration
Python programming
Generative AI
Data analysis
Git version control
Spark framework
Model evaluation
LANGUAGES
English Proficient, Isizulu Native
Accomplishments
- AI Immersion Heckathon Winner
- Lead CODE LIKE A GIRL Programme
- Vodacom Group Stuff Award Winner
Certification
- AI Immersion Heckathon - Advertising Campaign Effective
- Vodacom group Stuff Award Winner - Fraud Engine
Timeline
Data Science Specialist (CX & CVM)
Vodacom Blvd
02.2025 - Current
Junior Data Scientist (Fraud & CX)
Vodacom Blvd
02.2024 - 01.2025
Intern Big Data
Vodacom Blvd
02.2023 - 01.2024
B.Sc. Honours - Physics
University of the Witwatersrand
B.Sc. - Mathematics and Physics
University of the Witwatersrand