All Projects
InsightIQ
InsightIQ is a full-stack AI-powered analytics platform that allows users to upload CSV datasets and instantly generate visual insights, machine learning predictions, and natural language explanations. It combines traditional data science tools with modern LLM-based Q&A, enabling users to ask deep, conversational questions about their data and receive meaningful, contextual answers. With a beautiful, responsive interface and secure user authentication, InsightIQ transforms raw data into interactive and insightful dashboards.
MetaScan
MetaScan is a privacy-awareness tool that captures faces from live video streams using OpenCV, then conducts automated reverse image searches to investigate the digital traceability of those identities. It highlights the growing concerns around facial surveillance, showing how wearable or embedded devices can collect facial data without consent and potentially match it to online profiles. Built using Python, Selenium, and web scraping techniques, MetaScan explores ethical implications of computer vision and open-source intelligence (OSINT) in the AI era.
Speech-to-Math Calculator
This voice-powered calculator translates spoken mathematical expressions into symbolic equations and computes the result instantly. It supports operations like addition, subtraction, square roots, powers, and more. Built with Python and integrated with a Tkinter GUI, it combines speech recognition, natural language parsing, and text-to-speech to create an accessible tool for quick calculations without manual input.
Inventory Management Tool
A desktop-based inventory management application built using C++ and Qt for the frontend, and PostgreSQL for data storage. The system allows users to add, update, delete, and search product records efficiently. It replaces error-prone CSV-based management with robust SQL-driven workflows and includes features like product categorization, transaction history, and quantity tracking.
Smart PDF Q&A Tool
Smart PDF Q&A is a web application that allows users to upload PDF documents and ask natural language questions based on their content. It uses LLMs to chunk and embed document sections with semantic search (FAISS), enabling contextual understanding and interactive answers. The system supports summarization, question generation, and chat-based Q&A, making it ideal for reading academic papers, legal documents, or business reports efficiently.
Meso-American Pathfinders
Part of an archaeology-driven internship, this project explores historical trade routes in Mesoamerica by analyzing ancient place names. Using AI-powered language detection (TFLite), city names are classified by language (Spanish, English, or native), then geocoded and visualized in ArcGIS. The project merges linguistic analysis with geospatial mapping to reconstruct patterns of influence and migration in pre-Columbian Central America.