Data Science & Machine Learning Portfolio | Python, Scikit-learn & NLP | Samir Mohammed
Notebooks, pipelines, EgoQuest research, and applied ML.
Data science, machine learning, and AI portfolio
- Kaggle notebooks: XGBoost obesity-risk classification (~0.88 F1), bank churn prediction (~85% ROC-AUC)
- Hugging Face computer-vision tooling, ONNX inference pipelines, and open-source ML repos
- EgoQuest: conversational personality-typing research with RAG, ensemble ML, and transformer classifiers
- ETL and data pipelines: 35× throughput improvement (70 s → 2 s) for CRM record processing
- Domain analytics: astronomy pipelines, NGO impact metrics, and NLP for long-form narrative
- Tools: Python, Scikit-learn, XGBoost, LightGBM, PyTorch, Pandas, NumPy, SQL, Power BI
AI and NLP engineering
- Generative AI chatbots with GPT-3.5 and CRM integration for V75 Inc.
- NLP evaluation protocols and generative agent deployment
- RAG systems and retrieval-augmented generation for domain-specific Q&A
- Statistical analysis, EDA, and Tidyverse for Bureau of Statistics internship
- Computer vision: YOLOvX passenger and pothole detection (>90% mAP)