Luffy powers up with every contribution: Gear 2 at 400, Gear 3 at 800, Gear 4 at 1,200 and 1,600, Gear 5 at 2,000.
Get your own → luffygearup
I build data, machine learning and app projects, mostly in finance, risk and compliance. Each project below has a live demo and a write-up of how it was evaluated.
| Project | What it does | Stack |
|---|---|---|
| credit-risk-scorecard | Credit default model on 150,000 borrowers. WoE scorecard and LightGBM, SHAP explanations, expected loss in BHD. Demo | Python, LightGBM, SHAP, Streamlit |
| auditlens | Audit analytics combining 23 rule-based tests with unsupervised ML, evaluated against planted anomalies. Demo | Python, scikit-learn, Streamlit |
| cbb-compliance-assistant | RAG assistant for the Central Bank of Bahrain Rulebook with rule citations, refusal logic and a 30-question evaluation. Demo | Python, RAG, Streamlit |
| Akalt | Video-first food discovery app for restaurants in Bahrain. | Flutter, Firebase, Riverpod |
| Novessa-GymBro | Workout tracking with friend leaderboards that reward consistency. | Python |
Python, SQL, scikit-learn, LightGBM, SHAP, Streamlit, Flutter, Dart, Firebase, GitHub Actions

