BSc (Hons) Computing graduate specialising in Data Analytics, with a focus on applied machine learning, time-series forecasting, database systems, and backend software development.
Graduated in July 2026 with a degree awarded by the University of Greater Manchester and delivered through New York College Athens.
🏆 Student of the Year — BSc Computing (Data Analyst), Academic Year 2025–2026
I build reproducible data and software systems that combine rigorous analysis, machine-learning experimentation, database design, automated validation, and clear technical documentation.
- Applied machine learning for time-series forecasting and anomaly detection
- Renewable-energy analytics, model benchmarking, and residual diagnostics
- Leakage-aware evaluation and reproducible machine-learning experimentation
- Graph databases, Neo4j, Cypher, SQL, and data validation
- Privacy-preserving analytics and Differential Privacy
- Backend development with Java, Spring Boot, Spring Security, Django, and REST APIs
| Project | Area | What it demonstrates |
|---|---|---|
| WindPower Digital Twin | ML Research / Time Series | Leakage-safe wind-power forecasting, tabular benchmarks, graph-aware extensions, controlled sequence-model experiments, residual diagnostics, and a local Django artifact browser |
| DiffPriv-Gateway | Privacy-Aware Analytics | Differential Privacy middleware using Laplace and Gaussian mechanisms, privacy-budget controls, automated testing, CI, and GDPR-oriented documentation |
| Solar PV Anomaly Detection | Applied ML / Time Series | Unsupervised LSTM Autoencoder pipeline for detecting unusual photovoltaic generation behaviour from power and weather-sensor data |
| Cloud Inventory Graph Database | Database Engineering | Neo4j-first inventory database using Cypher, MariaDB support, Python validation pipelines, architecture diagrams, and GitHub Actions |
| Enterprise Task Management System | Software Engineering | Java 21 and Spring Boot application with Spring Security, RBAC, N-tier architecture, DTOs, CRUD workflows, and MySQL persistence |
A forecasting-first research repository for spatio-temporal wind-power modelling, evaluation, and diagnostics using the DaKS/Kassel synthetic renewable-power dataset.
The project includes:
- Strict raw-data validation and coverage-aware exploratory analysis
- Leakage-aware feature engineering and temporal train-validation-test separation
- Persistence, Linear Regression, Random Forest, XGBoost, and MLP benchmarks
- Controlled supplementary GRU, LSTM, TCN, and PatchTST-Lite sequence experiments
- Residual, operating-regime, and park-level diagnostic analysis
- Graph data contracts, graph-model input preparation, graph forecasting baselines, and spatial sensitivity analysis
- Validation-calibrated residual interpretation for future condition-awareness research
- Reproducible scripts, audit documentation, CI workflows, and repository safety checks
- A local Django interface for read-only inspection of curated research artifacts
The repository is intentionally presented as a forecasting and diagnostics research pipeline. It does not claim to be a deployed industrial digital twin, production forecasting service, fault-diagnosis system, or validated prognostics and health-management platform.
A Python-based privacy-preserving analytics middleware project designed for small and medium-sized organisations. It implements Laplace and Gaussian Differential Privacy mechanisms, privacy-budget awareness, clipping safeguards, synthetic datasets, automated testing, continuous integration, and GDPR-oriented documentation.
An applied machine-learning proof of concept using photovoltaic generation and weather-sensor data. It uses an unsupervised LSTM Autoencoder, time-series sequence generation, reconstruction-error thresholding, modular Python source code, and saved model and visual evaluation artifacts.
A NoSQL-first inventory-management database with Neo4j as its primary operational database. It includes Cypher constraints and operational queries, Neo4j Aura, MariaDB as an optional supporting relational layer, Python-based data validation, architecture diagrams, and automated repository checks.
A full-stack Java web application built with Java 21, Spring Boot 3.3, Spring Security 6, and MySQL. It demonstrates authentication, role-based access control, CRUD and filtering workflows, DTO and Repository patterns, and an N-tier Controller-Service-Repository architecture.
Languages: Python, Java, SQL, Cypher
Data Analysis: pandas, NumPy, Jupyter
Machine Learning: scikit-learn, XGBoost, PyTorch, TensorFlow, Keras
Databases: Neo4j, MySQL, MariaDB
Backend: Spring Boot, Spring Security, Django, REST APIs
Testing and Automation: pytest, Maven, GitHub Actions
Engineering: Git, CI workflows, reproducible project structures, security policies, and technical documentation
BSc (Hons) Computing — Data Analytics pathway
University of Greater Manchester
Delivered through New York College Athens
Graduated: July 2026
Academic distinction:
🏆 Student of the Year — BSc Computing (Data Analyst), Academic Year 2025–2026
My professional direction is focused on data analytics, applied machine learning, time-series forecasting, and data-intensive software systems.
I am particularly interested in work that combines reliable analysis, reproducible experimentation, renewable-energy or operational data, database engineering, and well-tested software development.