I started my career maintaining the networks that keep millions of people connected. For 13 years at Ericsson and Huawei, I was the person who looked at the data when something was wrong, found the pattern nobody else had spotted, and turned it into a decision.
Then I went further. I completed an MSc in Applied AI at Aston University, building an Ensemble ML classifier across five models with 99.12% accuracy on 5 million records. I hold certifications from Microsoft, Neo4j, and Google.
Thirteen years of evidence. Three certifications. Two master's degrees. One consistent thread: turning data into something useful.
Built a soft-voting Ensemble Classifier combining five machine learning models to detect four cardiovascular conditions (Heart Attack, Heart Failure, Angina, and Peripheral Artery Disease) from structured patient data.
Dataset: 5 million synthetically generated patient records with 72 clinical features including ECG parameters, blood pressure, cholesterol, troponin, ejection fraction, and WBC count. 80/20 train/test split.
Self-managed from research question through data generation, model development, evaluation, and written submission. Deployed as a real-time patient prediction tool.
| Classifier | Accuracy |
|---|---|
| Logistic Regression | 99.70% |
| Gradient Boosting | 99.17% |
| Ensemble Voting Classifier | 99.12% |
| Decision Tree | 85.03% |
| Random Forest | 79.08% |
| K-Nearest Neighbors | 68.52% |
The Ensemble Voting Classifier combines the predicted probabilities of all five base classifiers using soft voting, aggregating their individual strengths to produce a more robust and generalisable prediction than any single model alone.
Evaluated via accuracy, precision, recall, F1-score, confusion matrix, calibration curves, and precision-recall curves across all four cardiovascular disease classes.
Actively seeking roles across data analytics, BI, ML engineering, and digital project management. Open to all sectors. If you are a recruiter, hiring manager, or fellow data professional, I would welcome a conversation.
I bring 13 years of operational analytics, three professional certifications, two master's degrees, and a machine learning thesis achieving 99.12% accuracy on 5 million records. Available immediately. UK-based. Hybrid and remote considered.
Seeking positions in: