I am an accomplished Al Engineer with a proven track record of delivering innovative data-driven solutions. The drive in my career has been to promote innovation and adoption of advanced analytical solutions in various industries.
Overview
8
8
years of professional experience
Work History
Al Engineer & Innovation Leader
Deloitte Africa
01.2016 - Current
Comprehensive experience in building analytics and machine learning pipelines to solve complex business problems
Leader of the Highly Experimental Unit, a team focusing on bridging the gap between Al research and real-world problem-solving through high-risk experiments for clients
Considered as a subject matter expert in Al and Python in the Deloitte analytics community
Led workshops for internal and external clients, assisting them in brainstorming and planning how Generative Al should be used to evolve their respective businesses
Owner and maintainer of Deloitte's Al Python SDKs used for Anomaly Detection and Semi-supervised learning.
Education
MSc Computer Science (Distinction) -
University of Witwatersrand
01.2021
BEng Industrial Engineering (Distinction) -
University of Pretoria
01.2013
Skills
Innovation
Critical Thinking
Problem Solving
Machine Learning
Computer Vision
Reinforcement Learning
Semi-supervised Learning
NLP/NLG/LLMs
Spark/Pyspark
Pytorch, Langchain, Sklearn
Linux
CI/CD, Gitlab
Docker & Kubeflow
Azure, AWS, GCP
Stakeholder Management
Project Management
Collaboration Skills
Certification
AWS - Certified Cloud Practitioner
AWS - Machine Learning Speciality
University of Alberta - Reinforcement Learning Specialization
Recent Projects
Inventory Management with Reinforcement Learning, Design and implementation of a system that optimizes the scheduling, manufacture, and distribution of a single product. The solution saves the client $10k per month and enables them to simulate further scenarios to plan for the future development of their supply chain (PyTorch, Pygame).
Report Automation with NLG, Built a system that automates report writing of findings from fraud investigations at a major Telecommunications client. The tool saves the client thousands of hours per month (LLMs, Langchain, Azure OpenAI, PySpark).
Trained a model on a corpus of legal documents to act as a compliance assistant. The solution performs a variety of tasks, including vetting whether or not a contract meets legal requirements and answering questions about legislation (LLMs, Langchain, Azure OpenAI).
Predictive Maintenance, Supervised learning pipeline to monitor the operational performance of hundreds of trucks at a large mining company. The system is able to detect early warning signals of imminent machine failure, mitigating the risk of critical breakdowns, saving the client $100,000+ per month (PySpark, Scikit-Learn, Power BI).