Applied AI & Data Analytics Specialist experienced in building end-to-end Machine Learning pipelines, Star Schema BI modeling, RAG architectures, and uncertainty-aware decision-support systems. Currently a Data Analysis & AI Specialist Trainee under the Digital Pioneers Initiative (MCIT & Military Technical College).
- ๐
Microsoft Certified: Power BI Data Analyst Associate (PL-300)
Credential: 67EDAFF039951D31 - ๐ Google Data Analytics Professional Certificate (Coursera)
- ๐ง Supervised Machine Learning (Stanford Online & DeepLearning.AI)
- ๐ Machine Learning with Python (IBM)
- ๐ Professional Diploma in Applied AI & Data Analytics (MCIT)
- BI & Analytics: Power BI, DAX, Power Query, Star Schema, KPI Dashboards, Row-Level Security (RLS), Excel
- Programming & Frameworks: Python (Pandas, NumPy, Scikit-learn, PyTorch), SQL, FastAPI, Streamlit, Docker
- Machine Learning & Modeling: CatBoost, LightGBM, XGBoost, Conformalized Quantile Regression (CQR), SHAP, Temporal Validation
- Deep Learning & CV: PyTorch, EfficientNet, CNNs, OpenCV, CLAHE Preprocessing, Grad-CAM (XAI)
- Generative AI & LLMs: RAG Architectures, Vector Search, ChromaDB, LangChain, Arabic Embeddings
- Islamic Fatwa RAG System: Guardrailed RAG platform querying 139K+ Arabic legal & religious fatwas using ChromaDB, specialized Arabic vector embeddings, and Streamlit.
- ORCA: Supply Chain Decision Intelligence: Leakage-safe delay prediction pipeline (ROC-AUC: 0.833, PR-AUC: 0.270) with 95.1% CQR uncertainty coverage, deployed via FastAPI & Streamlit.
- Passive Cognitive Radar Counter-Stealth: Micro-Doppler PBR target classification (98.3% accuracy on 60 GHz data, 90.5% on stealth targets) with ~16ms latency using EfficientNet-B0.
- Diabetic Retinopathy Medical Vision System: Deep CNN classification with CLAHE enhancement for retinal disease detection.
-
Body Performance Analytics & ML Engine: Multi-model system combining Neural Networks and Regression (74.65% accuracy,
$R^2=0.79$ ) with an interactive Streamlit UI. - Academic Risk Analytics Engine: Risk-scoring and performance modeling across 2,400+ student academic records.
- ๐ผ LinkedIn: ahmed-ibrahim-zain
- โ๏ธ Email: ahmedzan91@gmail.com