Built an end-to-end Agentic RAG system that enhances traditional RAG with autonomous planning, tool use, and iterative retrieval. LLM agents refine queries, retrieve context from multiple sources, and dynamically decide retrieval strategies before generating responses.

Domain-specific RAG chatbot providing accurate, context-aware responses to nursing mothers' health and childcare questions. Retrieves from a curated medical knowledge base and generates safe, grounded responses while reducing hallucinations.

Comparative ML study on AI-generated vs human-written text across multiple large language models. Explores linguistic patterns, stylistic differences, and model-specific behaviours using NLP preprocessing and classification. Evaluates cross-model generalisation challenges in AI text detection.

RAG chatbot that lets users upload PDF documents and interact with their content through natural language. Extracts, chunks, and indexes PDF content into a vector database so the LLM can retrieve relevant context and generate accurate, grounded responses.

Conversational AI assistant for task-oriented interactions, enabling users to automate everyday actions and retrieve information through natural language commands. Integrates an LLM chat interface with modular tool and function handling for productivity and decision-making support.

ML classification model predicting breast cancer diagnosis (benign vs malignant) from clinical and diagnostic features. Covers data preprocessing, feature analysis, and supervised learning models to support early detection and improve diagnostic decision-making accuracy.

Supervised ML model predicting diabetes diagnosis from medical diagnostic features. Includes exploratory data analysis, feature scaling, and training multiple classification algorithms to identify the most effective model for early-stage detection.

Computer vision system for detecting and classifying vehicle colors from images. Applies image preprocessing and feature extraction to identify dominant color regions, enabling automated color recognition for intelligent traffic and surveillance applications.