An agentic AI-powered assistant that automates campus workflows through reasoning, retrieval, and real-world action.
Learn More
Built a CNN-based model with 92% accuracy for pneumonia detection from X-ray images, featuring a real-time Flask-powered user interface.
Learn More
Built a custom neural network library and applied it to model real-world NYC taxi trip durations with optimized performance.
Learn More
Built a TF-IDF-based search engine to retrieve the most relevant U.S. presidential inaugural speech using cosine similarity and inverted indexing.
Learn More
Implemented Decision Tree, Random Forest, and AdaBoost from scratch to classify Titanic survival outcomes with custom-built algorithms.
Learn More
Built an AI-powered job application assistant using Groq-hosted LLaMA 3.3-70B, LangChain, and Prompt Engineering to generate personalized cold emails for job applications.
Learn More
Developed a machine learning pipeline to predict house prices in Bengaluru, applying feature engineering, outlier removal, and regression modeling.
Learn More
Built a customizable two-pile Nim variation with AI powered by depth-limited MinMax and Alpha-Beta Pruning, supporting both standard and misère rules.
Learn More
Mav Chatbot – LLM-powered assistant for UTA with hybrid retrieval, smart crawling, and real-time deployment.
Learn More