CS5934: Capstone Project
Permanent URI for this collection
Browse
Recent Submissions
- Test4Test - AI Dashboard and Analysis for user testingOzoroski, Matthew; Li, Zekun; Yayloyan, Nelli; Linkous, Ryan (2026-08-15)Test4Test is a free user testing service for anyone with a website or app. For CS5934, we used AI to analyze usability test recording so users don't have to manually watch every recording, gain quick insights and see trends based on qualitative characteristics.
- ReadySetHome: A Personalized Home Purchase Planning PlatformHaddadin, Michael; Dudkowski, Joseph; Keane, Conor; Mahabal, Anshuman; Mishra, Pranav (2026-08)Buying a home is one of the largest financial decisions most people will make, yet the path to readiness is poorly understood by prospective buyers who are often years away from being able to purchase. Existing tools either assume a buyer is already mortgage-ready (listing sites, mortgage calculators) or offer generic budgeting advice disconnected from the goal of homeownership. ReadySetHome addresses this gap with a full-stack web application that builds a personalized, continuously updated roadmap toward home purchase readiness. Users provide a financial profile (income, savings, debts, and credit standing) along with a target home price, purchase area, and timeline; the platform calculates a down payment target, projects a debt-to-income ratio, and generates a sequenced set of milestones with a dynamically recalculated monthly savings requirement. A single readiness score (0-100) and a projected purchase date update automatically as users log financial progress, and a large-language-model assistant answers plan-specific questions grounded in the user's real financial data. The system integrates live mortgage rate data (Federal Reserve Economic Data), property tax data (U.S. Census Bureau), and real estate sale listings (RentCast) to anchor recommendations to current market conditions.
- AI Learn: Source-Grounded AI Course Support with Canvas IntegrationAssadzadeh, Arian; Aryal, Aashish; Huang, Bo-Chiao; Hsin, Liang-Wei; Patel, Harsh Gunvantbhai (Virginia Tech, 2026-08-14)AILearn is a full-stack, course-scoped AI assistant designed to help students find grounded answers from their course materials while giving instructors tools to manage and improve course support. Students can ask questions through streaming chat, continue conversations, inspect citations and source locations, save useful answers, provide feedback, and view course discussions. Instructors and teaching assistants can create course workspaces, import Canvas courses, rosters, announcements, deadlines, and materials, upload and reingest documents, images, spreadsheets, and media, and monitor processing status. A retrieval-augmented generation pipeline combines indexed course-material chunks with authorized read-only Canvas evidence and conversation context. The system validates citations and routes low-confidence or flagged answers to a staff review queue, where answers can be corrected and published to course discussions. Course analytics summarize usage, confidence, feedback, participation, repeated questions, and response performance. This submission contains the final AILearn presentation and demonstration video.
- Eleos Access: Diaspora Concierge HubZeno, Mary ; Vempuluru, Sruthi; Valaboju, Vrinda; Bangari, Neha (2025-08-16)Eleos Access is a multilingual, web-based platform developed for Eleos Groups, designed to support members of the Ethiopian diaspora in the U.S. and Europe by enabling them to plan and pre-pay for essential services before arriving in Ethiopia. The application provides booking modules for travel planning, property management, and business support - each of which are reviewed and processed by Eleos Groups. A travel estimator powered by the Mistral AI model allows users to calculate projected travel costs, while integrated PayPal payment processing ensures secure online transactions for service requests. Users can register for personal accounts to submit their service requests and track payments, while administrators can oversee request payments and manage property listings. The system is built with a React.js frontend using Vite, Shadcn, and Tailwind CSS, and a backend powered by FastAPI and Uvicorn. Database management, user authentication, and email notifications are done through Supabase.
- CS5934 Capstone Project: LeafRXHart, Sabrina; Park, Brian; Biju, Rayhan (2025-08-15)
- RAG Chatbot For Creator InsightsWakeley, Christopher; Hu, Deyu; Folomeev, Valery; Puppala, Sai (2025-08-15)
- WagelboxPlusPlusAlexander Keehan; Harold Velasquez; Nic; Daniel Montoya; Ahmad Kassem (2025-08)Agile development with client to improve their existing email marketing website by adding AI generation functionalities to email campaign creation to improve existing user experiences. This included, adding image recognition to have more relevant images recommended to a user, image generation to generate more relevant images, and email campaign generation to AI generate an entire email campaign for a more professional result. Project also included work on their website by modernizing, fixing bugs, improving existing features, and making it mobile responsive.
- Financial Data Dashboard for Old Town Curated/Alps CFOHoang, Nhan; Wynn, Benjamin; Scott, Layla; Dugan, Parker (2025-08-12)
- Melio ModelsKirillov, Maksim; Schermerhorn, Erica; Brassard, Benjamin; Hyde, Hunter (2025-08-11)A symbiotic object-recognition site - where you and object recognition models learn together. Melio Models harnesses the natural human desire to learn, creating a symbiotic relationship between knowledge seekers and object recognition models. Like curious minds, these models improve with every new piece of input. By providing a platform where individuals and models share knowledge, both sides benefit. When a human’s knowledge falls short, the model can try to fill in the gaps in the Object Identifier. When the human wants to learn and test their skills, they can play the Learning Game. When the model needs improvement, the human can guide it through feedback. And when the model needs to expand the items it can recognize — or when an entirely new model is needed — that can be done in the Model Explorer.
- Minecraft Final Presentation Demo - Group 10Shah, Faraz Ulhaq ; Kumari, Sanjna; Assadi, Taeesh Azal (2025-05-04)
- Career Compass: an AI career advisorKarne, Anvita Sanjay; Chouthai, Atharva Bipin; Kumar, Drushya Chikkaballapur; Gutgesell, Julia Marie; Maddipati, Mahita; Belgundi, Rajat Ratnadeep; Deshmukh, Shubham Laxmikant; Kasera, Siddhi (2025-05-05)Career Compass is an AI-driven career guidance platform that combines Retrieval-Augmented Generation (RAG), structured labor data from O*NET and BLS, and large language models (LLaMA 3.3) to deliver personalized, explainable career insights. The platform offers a real-time conversational chatbot and an embedded Career Insights Dashboard built in Tableau, enabling users to explore occupations, required skills, education, and job outlooks interactively. Designed for accessibility and scalability, it includes a semantic vector search engine, persistent conversational memory, and structured fallback logic to ensure accurate responses. The data visualization component translates complex employment and wage statistics into narrative-driven dashboards and interactive data stories, making labor trends more understandable for users without technical backgrounds. The application is deployed on AWS EC2 with a Next.js frontend and FastAPI backend.
- U-Pass ManagerKamath, Nidhi Govindraya; Mahato, Kunal; Nagar, Rajat; Kulkarni, Rashmi Dattatraya; Tirkey, Manim; Huang, Cheng-Chen (2025-05)The U-Pass Manager is a cloud-based web application designed to automate and streamline the distribution of university transit passes, replacing outdated, manual Excel-based workflows with a scalable, secure, and real-time platform. Developed as part of a Capstone Project, at Virginia Tech, the system features robust role-based access controls for administrators and distributors, integrated NFC scanning for quick card assignment, AWS-powered backend services for data processing and communication, and a detailed analytics dashboard to support informed decision-making. By eliminating inefficiencies, the application enhances the accuracy, transparency, and scalability of public transit access for students, while enabling seamless coordination with external organizations such as WMATA.
- PitDash: Live Surgery DashboardLowry, Keith; Hoang, Brandon; Pradhan, Hansa; Kulkarni, Yash; Yeboah, Tesha; Panigrahi, Asmi; Bhavsar, Revati; Manavalan, Sudarshan (2025-05-09)
- HomiezPark, Esther; Rama, Shalini; Gowrishankar, Vasundhara; Somashekar, Shanmuganathan (2025-05)Homiez is a web application that empowers property buyers and renters to stay updated on real-time price changes in the real estate market. Unlike traditional real estate platforms that prioritize advertisers, Homiez puts users first, offering a focused, distraction-free experience to track property price drops and increases. By simplifying how users monitor property trends, Homiez enables smarter, faster, and more informed buying and renting decisions.
- GrantVaultCui, Timothy (Tikki); Zimmerman, Ian; Naseh, Harris; Dhongade, Shardul (2025-05)
- Roomly ApplicationPatel, Shalini; Shearon, Kelsey; Bansal, Ishi; Ganta, Kalyani; Anandarajan, Shareeya (2025-05-07)
- Greek RightUlloa, Simon; Eichensehr, Vanessa (2025-05)GreekRight is an application for fraternity and sorority management. Our application allows student leaders to manage their organization and allows members to access the information that they need. Our current solution has three major components: a calendar system to keep members informed of events, a shifts system to coordinate volunteers, and a messaging system for streamlined communication. We plan to continue to add features and allow each organization to choose and customize the features that best suit their needs.
- EnQueueBone, Spencer; Goodman, Patrick; Yi, Branden; Ogura, Griffin; Kanjoor, Ajay (2025-05)This project presents a web-based academic support platform integrated with Canvas via LTI and API, built using FastAPI and PostgreSQL. It provides a unified interface for managing student help queues, TA schedules, course rosters, assignments, and FAQs. With features like role-based access, TA location tracking, in-queue student games, and TA rating systems, the platform aims to streamline academic support in large university courses. By combining structured course data with real-time interactions, it improves responsiveness, reduces redundancy, and enhances the learning experience for students and teaching assistants alike.
- CS5934 - NavagantAIDewey, Patrick; Gomez, Aaron; Jaimes, Alex; Kamalesh, Akhil (2025-05)NavagantAI is a prospecting platform built for our client Navagant, a Richmond-based mergers and acquisitions advisory company. The core goal of this platform is to assist associates at Navagant in finding new clients through the use of historical outreach data and a machine learning model. NavagantAI was architected to ensure robustness, performance, and security, with a user-first design. The core feature of our application is company search, which finds companies based on various parameters (sector, keywords, number of employees, revenue, etc.) and ranks the results using a similarity score assigned by our model. Our search system is considerably faster than competing systems, retrieving results in just a few minutes—compared to the hours taken by similar commercially available systems. Users can view search results in tabular form from a web browser or download the results for use outside of the application. Searches and results are also saved, allowing users to revisit previous search queries and results. The speed and intuitive nature of NavagantAI position it as a powerful tool for prospecting, with the potential to redefine how Navagant finds new clients.
- TutorTech: An Easier Way to Find Tutors and TuteesKolli, Atin; Elavarthi, Bhargava; Morris, Jerry; Nair, Krishna; Kersulis, Tomas (2025-05-07)