Data Scientist & Aspiring Applied AI Engineer — Computer Science & Statistics graduate (Helwan University) focused on NLP and LLM-powered systems.
- 🧠 Building pipelines that turn unstructured, informal text — including Arabic dialects — into structured, reliable outputs
- 📊 Solid foundation in statistics and ML: classification, clustering, model evaluation, BI reporting
- ⚙️ Design-first mindset: typed contracts, validation, and traceability, not just model accuracy
- 📚 Continuing to build depth in Applied AI, NLP, and LLM applications through hands-on projects
Interested in opportunities across Applied AI, NLP Engineering, and Machine Learning, Data Science where I can contribute to building reliable, language-centric AI systems.
Applied AI & Generative AI • Natural Language Processing (NLP) • Machine Learning • Data Science • Data Visualization & Business Intelligence
| 🤖 Machine Learning | 🗣️ NLP & Generative AI | 📊 Data & BI |
|
Classification & Clustering Model Evaluation Hyperparameter Tuning Feature Engineering |
Arabic NLP LLM-Based Applications Prompt Engineering Entity Resolution & Fuzzy Matching |
Data Cleaning & EDA Power BI / Tableau SQL Server BI Stack (SSIS/SSAS/SSRS) Data Warehousing |
Graduation Project — Excellent Grade
Summary: Converts unstructured Egyptian Arabic food orders — informal dialects, non-standard expressions — into accurate calorie estimates via a modular five-stage pipeline (normalization → LLM extraction → entity resolution → gram-weight estimation), built on strict typed contracts.
Highlights:
- Confidence-tiered entity resolution: exact matching, stemming, and fuzzy matching with field-specific thresholds
- Automatic correction of upstream LLM extraction errors via density-based lookup
- Decimal-only arithmetic with full audit trails for precision and traceability
Tech Stack: Python Pydantic Hugging Face Transformers Qwen2.5-7B RapidFuzz JSON
