Abdur Rahman Riad

An advanced RAG AI Assistant built with Streamlit, LangChain, ChromaDB, HuggingFace Embeddings, and Google Gemini 2.5 Flash. Allows users to upload PDF documents, automatically extract and process the content using OCR technology, convert it into searchable vector embeddings, and then ask natural language questions about the document. The system provides accurate, context-aware answers based strictly on the uploaded document content.

A bilingual Q&A chatbot designed to answer questions about Bangladesh in both English and Bangla. Created using GenAI LLM with Langchain and Gemini API. The chatbot provides accurate, context-aware answers to help users learn about Bangladesh's culture, history, geography, and current affairs. Features natural language processing in both languages for seamless user interaction.

An end-to-end Retrieval-Augmented Generation (RAG) application built with Streamlit, LangChain, ChromaDB, HuggingFace Embeddings, and Google Gemini 2.5 Flash. Users can enter any website URL, and the application automatically scrapes the content, converts it into searchable vector embeddings, and then enables natural language questions about that website's content.

A comprehensive data analysis and visualization project built with Python to explore, understand, and visualize key socio-economic indicators across the world. Analyzes Global Datasets on GDP, population, education, CO2 emissions, inflation, unemployment, and population growth. Features geospatial and 3D visualizations using Plotly and Folium to identify relationships and disparities between countries' economic indicators.

A complete data analysis and forecasting project built using Python to explore, visualize, and predict crime trends across different divisions and years in Bangladesh. This project combines Data Preprocessing, Statistical Modeling, Machine Learning, and Interactive Visualizations to provide data-driven insights. Uses ARIMA, LSTMs, Linear Regression, and Random Forest for prediction. Includes Matplotlib, Seaborn, and Plotly visualizations.

A comprehensive Machine Learning and Data Analysis project developed with Python to identify and evaluate key factors contributing to student depression. Examines stress, sleep, social life, and academic workload to identify patterns and predictive factors. Uses Data Preprocessing, Exploratory Data Analysis, and ML models (Logistic Regression, Random Forest, SVM) to predict depression levels with 90%+ accuracy. Provides insights with interactive visualizations.

A complete Data Analysis and Machine Learning project built with Python to explore, clean, and predict passenger survival outcomes from the Titanic dataset. Covers the complete data science pipeline including Data Preprocessing, Visualization, Feature Engineering, Model Building, and Evaluation. Implements Logistic Regression, Decision Tree, and Random Forest classifiers achieving 87% accuracy. Provides interactive charts for demographic insights and prediction performance.