Digital Verified Driver
Graduation project: edge AI against disability-vehicle misuse and theft, plus cabin safety for ride-hailing and schools — Raspberry Pi, YOLOv5, InsightFace, FastAPI, MongoDB.
Graduation project: edge AI against disability-vehicle misuse and theft, plus cabin safety for ride-hailing and schools — Raspberry Pi, YOLOv5, InsightFace, FastAPI, MongoDB.
End-to-end RAG system over ~80K Arabic passages in 10 knowledge categories. Every answer cites book, author, and page; the assistant refuses when no relevant source is retrieved. FastAPI, PostgreSQL + pgvector, Celery + Redis.
End-to-End AI platform for classifying blood cell images (+95% accuracy): 17K+ images (8 classes), custom PyTorch CNN, Flask web app, Docker deploy.
Web app predicts Boston house prices with Decision Tree ML. Flask backend, REST API, live demo inputs & full data analysis. Production model, modern UI, 78.8% R² accuracy.
NLP app classifies news headlines (SPORTS, CRIME, EDUCATION, COMEDY) with 84% accuracy. Multinomial Naive Bayes + CountVectorizer, Gradio UI, Hugging Face Spaces deploy.
Modern Flask app to predict weekly Uber rides based on simple inputs. Styled with an Uber-like theme, production-ready setup with scikit-learn model, and deployed on Railway.
Machine learning project that analyzes and predicts flight delays based on carrier information, airport details, and flight statistics. Optimized Random Forest model achieves ~0.99 R² score with ~1.9 MAE.
AI-powered travel planning application built with Flask. Uses OpenAI GPT-4 for intelligent travel recommendations, real-time weather data, cost calculations, and search integration for hotels, attractions, and restaurants.
Intelligent chatbot application built with Flask and OpenAI. Features conversational AI capabilities, context-aware responses, and a modern web interface for seamless user interactions.
Text classification project to predict Positive or Negative sentiment from Amazon product reviews. TF-IDF + Logistic Regression, NLTK preprocessing, and a Flask web app with dark modern UI.
Screen and rank resumes against job descriptions using Sentence Transformers, cosine similarity, and a Flask web UI. Upload PDF/DOCX/TXT resume and see top job matches with scores.
Extractive question answering system using Transformer models fine-tuned on the Stanford Question Answering Dataset (SQuAD), with a Flask web interface to ask questions and get answers from context.
Flask service exposing a logistic-regression model trained on match-level statistics to estimate a team's chance of victory, with a reproducible training pipeline, JSON API, and an interactive dashboard.