FINTECH · ACCOUNTING · MACHINE LEARNING · DIGITAL ASSETS
FinTech builder and researcher creating applied systems across quantitative finance, accounting analytics, machine learning, AI, blockchain and digital assets.
ABOUT
I am a financial technology builder and researcher pursuing an M.S. in Financial Technology at the University of Central Florida, with a B.Sc. (Hons) in Applied Accounting from Oxford Brookes University. My background gives me a practical view of both sides of financial innovation: how financial records, controls, budgets and business operations work, and how data, machine learning and software can improve them.
My interests span quantitative finance, accounting analytics, digital assets, blockchain security and applied machine learning. I enjoy taking a financial problem from concept to implementation—designing the data pipeline, selecting and evaluating models, and turning the result into a usable web application or research benchmark.
Alongside technical work, I have entrepreneurial operating experience as Co-Founder of Charles Beatina PBC, a meat-processing startup. Managing purchasing, budgeting, supplier decisions, cash flow and day-to-day operations strengthened my interest in building technology around real business and financial problems.
Python · PyTorch · LightGBM · XGBoost · Graph Neural Networks · SQL · Solidity
SELECTED WORK
An applied quantitative-finance platform for exploring portfolio downside exposure and hedging decisions. Hedge Master brings market-data analysis and risk calculations into a simple web workflow and produces automated summary reports for communicating the analysis.
Visit project ↗A probabilistic market analytics application that uses LightGBM and engineered distribution, technical and market-context features to classify equity return environments into Large Down, Small Down, Neutral, Small Up and Large Up regimes. The interface emphasizes probabilities so users can see model uncertainty rather than only a single label.
Visit PMRC ↗An applied accounting-AI platform built around transaction-ledger risk analysis. AuditShield AI explores automated identification of normal activity, accounting errors and potentially fraudulent journal-entry patterns, connecting accounting-domain knowledge with supervised and relational machine-learning approaches.
Visit AuditShield ↗A multimodal financial-risk research project that combines structured corporate financial information with narrative disclosure text to estimate forward corporate distress risk. The project explores how quantitative fundamentals and language-derived information can complement one another in predictive modeling.
View research project ↗Research-oriented deep learning using Bitcoin on-chain transaction-network chainlets as structured signals for price forecasting. The project explores multilayer perceptron and graph-based architectures to connect blockchain transaction structure with subsequent market behavior.
An open-access synthetic ERP accounting benchmark designed for machine-learning experiments on financial risk. The dataset supports research on normal transactions, accounting errors, fraud patterns, transaction gaps, approvals, user behavior and verification trails.
Explore dataset ↗CAPABILITIES
Accounting · ERP & ledger analytics · reconciliations · budgeting · financial modeling · quantitative finance · digital assets · blockchain · financial risk analysis
Python · HTML · JavaScript · R / RStudio · SQL · Solidity · Full-Stack Development
Machine Learning · Large Language Models (LLMs) · Artificial Intelligence · scikit-learn · TensorFlow · PyTorch · LightGBM · XGBoost · Graph Neural Networks
ENTREPRENEURSHIP
Co-founded and operated a meat-processing startup, combining hands-on business operations with financial management. My responsibilities included operational budgeting, supplier evaluation, purchasing, purchase orders, commercial disbursements, budget-variance reviews and monitoring cash flow and margins.
This experience shaped how I approach FinTech: technology should solve operational problems, improve financial visibility and support better decisions—not exist only as a model or prototype.
RESEARCH
Comparative research on structured, temporal and relational learning for automated transaction validation and accounting anomaly detection.
XGBoost · Random Forest · LSTM Autoencoder · GraphSAGE · Edge-aware GraphSAGE
Benchmarking machine learning, opcode representations and graph neural networks across Ethereum smart-contract vulnerability datasets.
Ethereum · Opcode Analysis · GNNs · Smart Contract Security
ACADEMIC REFERENCES
Assistant Professor of Computer Science, FinTech and Digital Forensics. Academic reference connected to my graduate FinTech and technical work.
Lecturer, National University of Science and Technology (NUST), Zimbabwe.
CONTACT
For professional opportunities, FinTech projects, research collaboration or conversations about financial AI and blockchain.