Cornell Bowers College of Computing and Information Science

BOOM 2025 Projects

Check out the Projects for BOOM 2025

Thank you to everyone who submitted all the amazing project ideas!

Learning in Queuing Systems with Buffers

Jeronimo Martin Duque

In our project, we study packet buildup in a network in which routers compete to get their packets processed by servers. We model and analyze this system mathematically to derive conditions that ensure there is no significant buildup. Our key contribution is the presence of buffers at the servers, which alters the behavior of the routers. This is complemented by simulations of the network, allowing us to better understand the dynamics of such systems.

Locket Widget

An Le

Locket is a widget that shows live photos from your best friends, right to your Home Screen. You and your best friends will see new pictures from each other every time you unlock your phone. It’s a little glimpse of what everyone’s up to throughout the day. We are currently #15 in Social Networking on the app store with 393.7K Ratings.

iName - Name Recognition & Pronunciation App

Shreayaa Nadagudy Srinivasan, Ariel Zou, Chaowei Xiao, Qiqiu Qian

iName is a real-time, culturally aware name recognition and pronunciation system that fosters inclusivity. Users can record name pronunciations, explore cultural backgrounds, and integrate with smart devices to ensure correct name usage in social and professional settings. iName addresses the exclusion faced by non-English speakers and promotes cultural identity through features like group-based name learning and an interactive world map.

Bump

Ilincev Simon, Vagish Eashan

Bump is an iOS app that seamlessly records all your in-person interactions with friends and provides you with a real-time directory of those around you. We turn passing moments with pals into lasting memories, and grant the ability to stay connected to any social scene from wherever you may be.

Uplift - Cornell Fitness

Caitlyn Jin, Andrew Cheung, Joshua Dirga, Stefanie Rivera-Osorio, William Lee

Cornell lacks a centralized platform for fitness activities and resources. With the demanding academic environment, students often overlook prioritizing their physical and mental health, leading to a lack of motivation to stay active. Uplift provides a personalized fitness platform that seamlessly integrates into the daily routines of gym-goers, offering access to up-to-date information on gyms, facilities, and fitness classes.

Resell

Cindy Liang, Ashley Herrera, Justin Guo, Peter Bidoshi, Shrayes Gunna

Cornell students seek to buy and sell pre-owned items for college life. Current methods involve the annual dump and run, Facebook marketplace, or class/organization group chats. Issues with these current models involve inaccessibility, lack of user credibility, and disorganization. Resell is a mobile application that aims to be the central hub for Cornell students to sustainably and reliably purchase and sell pre-owned goods.

Glimpse

Larry Tao, Ethan Kim, Hannah Whang, Nicole Yan

Glimpse combines Internet of Things (IoT) technology with point-of-sale (POS) systems to offer real-time insights on restaurant food waste. We hope to equip small restaurant businesses with tailored and actionable insights for optimizing their supply chain, while also facilitating sustainable food waste disposal practices. We do this through a physical device that collects real-time data, and an algorithm that generates visualizations and insights on a user dashboard.

Carriage Renewed - a Product by DTI

Darin Lee, Desmond Atikpui, Matthias Choi, Ryan Qiu, Selena Liu

Carriage is a ride share app that allows those with disabilities to easily travel across campus. Partnered directly with CULift (Cornell's paratransit service), Carriage is dedicated to making Cornell accessible for all. Carriage targets students of all backgrounds, ensuring that injuries or long-term physical illnesses do not hinder students from attending classes and activities.

Subito Music Education Platform

Alexander Noviello

Subito is a music education platform with advanced drill generation, practice tracking, sheet music vector search, sheet music generative AI, and optical music recognition. We've developed our own linear sheet music encoding, with which we've built a variety of novel AI and non-AI technology for instrument practice, bundled into one practice tool for students and teachers. Our mission is to make instrumental music education more efficient, rewarding, and accessible for all.

CritterEvo

Tailai Ying

CritterEvo is an artificial life simulator used to model evolutionary dynamics within an artificial ecosystem. Players configure parameters such as world size, resource distribution, and mutation rates to customize a procedurally generated world populated by critters whose main goal is to survive and reproduce. Using genetic algorithms and natural selection, these critters adapt over generations, developing unique behavior and trait sets optimized for survival.

Trigger Happy

Pedro Pontes Garcia, Elaine Ran, Linda Hu, Luke Leh

Trigger Happy is a fast-paced last-man-standing mobile game where 3-6 players aim to defeat their opponents using quick thinking and strategy.

It's the 1920s in Chicago, and you find yourselves in the backroom of a smoky, seedy speakeasy. Settle scores in a high stakes game of cards, where you bluff, steal and shoot your way through a rapid-fire showdown. With bullets scarce and trust even scarcer, if you don't play your cards right, you'll be left holding a dead man's hand!

SeeThru

Bennett Santora, Abhimanyu Kumar

SeeThru is the first social finance platform that lets creators verify their earnings and build trust—on their own terms. SeeThru solves the problem of fake gurus lying without consequence by allowing users to connect financial accounts and verify earnings. Honest creators gain trust and grow, while users find verified influencers and track their own financial milestones.

A Journey Towards Sound and Complete LLM Planning

Daniel Cao

LLMs have demonstrated significant progress in code generation and debugging for hard reasoning tasks but how well they work on AI planning/search? We show they can do amazingly when queried correctly in our framework: AutoToS. It guides the LLM step by step towards the generation of sound and complete search components through feedback from both generic and domain specific unit tests. We achieve 100% accuracy with minimal feedback iterations with LLMs of various sizes on all evaluated domains.

Navi- Campus navigation

Anvi Savant, Angelina Chen, Christine Tao, Mihili Herath, Nicole Qiu

Navi is more than just campus navigation– it's your personalized guide to Cornell and Ithaca. Not only does it get you from point A to point B, but it also helps you discover everything in between. Tailored specifically for Cornell students and Ithacans, our app offers a local touch that you won't find elsewhere.

We're elevating Navi with notifications, trip sharing, user route reporting, and a comprehensive campus information ecosystem – on both iOS and Android, wrapped in a bold new identity.

Le Goosery Store

Liam Giraldo, Yizhen Wu, Zach Seidner

Le Goosery Store is a tower defense video game designed for both mobile and desktop devices. In this game, you play as a grocery store owner, who is facing trouble with thieving raccoons! To stop the thieves from taking your groceries, you must hire staff throughout your grocery store to prevent the raccoons from stealing.

Le Petit Raccoon

Amy Li, Angelina Chen

Le Petit Raccoon is a two-phase video game combining stealth and cooking featuring Rocko, the master chef raccoon opening up their very own restaurant. However, Rocko has forgotten to purchase ingredients! In order to keep the restaurant open, the player must aid Rocko in pilfering ingredients from a sumptuous yet fortified grocery store, cooking delicious meals, and feeding the hungry!

Kernel Functions

Yacqub Mohamed

There is tremendous interest in running multiple applications on memory-constrained devices for industries such as agriculture and supply chain. But running multiple applications requires isolation, which only a traditional OS provides in a memory-wasteful way. We designed Kernel Functions, an operating system architecture where system calls and page fault handling are done on a single small kernel stack, orthogonal to a traditional OS, while still providing traditional OS services.

Score - Cornell Sports

Yucheng Shu, Adelynn Wu, Helen Bian, Jay Zheng, Luke Stewart

Looking for a seamless way to follow Cornell sports? Score is your all-in-one a sports discovery hub for tracking upcoming schedules, live scores, and past records. Stay connected to Big Red anywhere, anytime!

OOI: A Search Engine for Project Ideas

Yuvraj Sarda

I've curated 15,000+ project ideas from across the internet, used LLMs to refine and evaluate them, and integrated it all together into a search engine. Next time you need an idea, you can just go here, say what you're looking for, and the search engine will show you the best ideas there are!

Deploying LLMs on Neural Processing Units

Anthony Fei

As LLM models continue to be popular, a particular open area of research is hardware acceleration and how to make computations faster. During my summer, I worked on developing the software to support LLM deployment on a new hardware unit called the neural processing unit (NPU), which was previously not possible. My work addresses a critical research gap in hardware acceleration, paving the way for faster, more accessible AI applications on mobile devices.

Joblink

Ahmed Abdulla, Farhan Mashrur

We’re revolutionizing the way candidates track their job applications by using AI to automate the entire process. Our platform streamlines application tracking by using a custom NLP classifier to parse recruiter emails, identify updates, and organize them into a clear, real-time timeline. No more missed emails, manual spreadsheets, or status confusion—just effortless, accurate tracking from outreach to offer.

CUApts: Smarter Housing for Cornell students

Celline Lee, Jaeyoung Lee

CUApts is a centralized platform built to simplify Ithaca’s complicated student housing process. Designed for Cornell students, it replaces scattered spreadsheets, word-of-mouth tips, and endless group chats with one clean, reliable website. With smart search filters, direct landlord messaging, and a peer-driven advice page, CUApts streamlines students' housing journey from search to lease.

Connect & Thrive: Smart Social Meetup Planning

Sumit Gawali, Aditya Vinodh, Isha Madhurendra, Peipei Duan

Graduate students often struggle to coordinate social meetups while balancing academic workloads. Connect & Thrive is a UX-driven solution that streamlines event discovery, scheduling, and spontaneous meetups. It personalizes recommendations based on user preferences, integrates academic and social calendars, and fosters community engagement through real-time meetup suggestions, enhancing student well-being and social life.

Schedula

Mike Zheng

Help Cornell students to coordinate and take courses together. Students share their course schedule on schedula and add friends and view each other's schedule.

ChitterChatter: AI Language Practice Partner

Jadon Geathers

ChitterChatter is an AI-powered speaking partner that helps students strengthen their language skills through natural, voice-based conversations. Fully integrated with classroom curricula, it provides practice opportunities precisely aligned with instructors' lesson plans. Students receive focused feedback after each conversation, building confidence and proficiency through low-stakes practice while addressing the limitations of traditional language education at scale.

Automated Coagulant Dosing Controller

Anjali Asthagiri, Claire Chiu, Claire Wang, Lauren Hsu

1 in 3 people globally do not have access to safe drinking water. This lack of access in small and rural communities is often due to the financial and technical barriers associated with chemical dosing. Our automated chemical dosing system reliably adapts to raw water conditions by integrating novel mechanistic models, sensors, and remote access capabilities — ensuring sustainable and effective drinking water treatment that meets regulatory standards.

Robot Pet Dog Benjavin

Kevin Lee, Benjamin Reyes

Robot dog "Benjavin" is a robot pet dog that follows you wherever you go. It uses an esp32-cam to detect your hand and if your hand is within the range where the cam module can detect it using a YOLO model, the robot will follow you just like a pet. The quadrabot has eight motors, two for each leg joint, enabling precise movement. Its computer vision system allows Benjavin to navigate and track its owner with remarkable accuracy.

Perceptions of AI Chatbots as Language Tutors

Daria Badger, Ava Friesen, Baihe Peng, Rachel Wang, Subhadra Das

This study evaluates student perceptions of the effectiveness of AI chatbot language tutors as compared to human tutors, specifically in providing feedback on student-written sentences. We examine which aspects of feedback students find most beneficial or limiting in each approach, as well as students' preferences between tutors. Through this, we seek to determine how language learning tools can be enhanced to create more accessible and beneficial tutoring experiences for students.

AI Environmental Impact

Nuo Cen, Baihe Peng, Steven Zhou

Our project is an interactive, data-driven article that explores AI’s environmental impact, focusing on the energy and water consumption involved in generating AI outputs. It features textual explanations, data visualizations, and an interactive calculator that lets users estimate their AI-related resource use. Drawing on public datasets and government reports, the project aims to foster transparency about AI’s hidden environmental costs and empower informed decision-making.

SeeRound

Naijei Jiang, Gene Wicaksono

SeeRound is a smart wearable designed for the visually impaired, using cameras and spatial audio to enhance indoor navigation. It detects nearby objects and people, translating their locations into directional sound cues for real-time awareness. Unlike canes, which struggle with overhead and dynamic obstacles, SeeRound provides a seamless, hands-free solution for safer movement indoors.

Carrying Capacity Early Warning System

Selena Zheng, Daniel Baldeo-Thorne, Nihaal Konda, Srija Ghosh, Tim Adeyemi

Cornell Hack4Impact is creating a Carrying Capacity Early Warning System for GreenZone, a non-profit based in Mongolia, to combat degradation from overgrazing and climate change within their rangelands. Our system provides data on rangeland health, so herders, policymakers, and analysts make informed decisions to prevent overexploitation. This approach ensures sustainable rangeland management, reduces desertification risks, and supports the livelihoods of communities dependent on these lands.

Power Assist Cart

Lili Mkrtchyan, Arshia Agrawal, Gabriella Best, Ignacio Estrada

A power-assisted housekeeping cart inspired by Cornell’s Statler Hotel. Designed for better ergonomics, reduced strain, and improved staff experience—while exploring human-robot interaction in hospitality settings.

DGA Studios Presents: Grub Bug

Phoebe An, Elmo Tang, Jacky Pan, Jacob Seto

Cornell’s Development in Games Association presents Grub Bug! Play as Cook Roach, a desperate yet determined father who’s taken to the streets to make ends meet the only way he knows how: selling grub. With no time to wait for a food vending license and cops closing in, you’ll cook, deliver, and drift through danger to outmaneuver every roadblock. Manage your oil, elude patrols, and dodge pedestrians—every mistake escalates the chase. Can you feed all against the odds?

Cornell Rocketry Software

Cameron Goddard, Arya Bhandari, Benjamin Zou, Max Larsson, Ronit Totlani

Cornell Rocketry is a 50+ student project team that is developing a high-powered, hybrid-propellant rocket from scratch. This year, we have been developing a cohesive software system to fuel, launch, and recover the rocket, while getting back essential telemetry every step of the way. This year represents a massive step up in the volume, complexity, and criticality of our software. Our work spans embedded systems, networking software, systems programming, UI design, and app development!

HireVision: Empower Talent

Pradhi Pakkerakari, Aryan Kumar

A step toward solving the age-old question of talent acquisition: "How do we hire and keep talent?" HireVision is the first video interview platform of its kind to offer an inclusive, scalable, and purely customaizable video interview experience that advocates for the candidate by providing the same immediate, downloadable AI-based summary and projection to both the candidate and the recruiter. Using methods derived from personality psychology research, HireVision ensures candidates and employers find their best respective culture and vibe fits, while also reducing the time to hire for companies.

Dataset Distillation for Foundation Models

Zhixin Sun

Our project utilized a self-supervised model to distill a smaller dataset. The smaller dataset can be used to pre-training other foundation models , saving resources and boosting performance.

Smart Collaboration Assistant

Liangcheng Dong, Guanlin Chen, Zhongyuan Wu

Our product is designed to improve team productivity and collaboration, not just attending meetings. It can join both online and offline meetings. During the meeting the assistant provides insights, mediates conflicts, and simulates the viewpoints of absent members. After the meeting the assistant summarizes the meeting and tracks and monitors the progress of follow-up tasks and provides real-time advice to team members.

CiteGeist: Automated Related Work Generation

Claas Beger, Carl-Leander Henneking

We present Citegeist: An application pipeline using dynamic Retrieval Augmented Generation (RAG) on the arXiv Corpus to generate a related work section and other citation-backed outputs. For this purpose, we employ a mixture of embedding-based similarity matching, summarization, and multi-stage filtering. To adapt to the continuous growth of the document base, we also present an optimized way of incorporating new and modified papers.

Clutch Coach

Dylan Price

Clutch Coach helps student-athletes up their game with personalized, in-person coaching and training through a peer-to-peer platform. Our platform provides youth and high school athletes with the unique opportunity to train directly with college and professional athletes, gaining insights, skills and feedback tailored to their specific needs.

Catacombs

Lawrence Daniels, Allen Chen, Happy Li, Joaquin Rivera

Catacombs is a 2D action strategy game made for PC.  Seize control of ancient Egyptian enemies while on a journey to the afterlife!

ACT Mobile Safety Deck

Aiden Montesinos, Arianna Hsu, Brian Sa, Jessica Andrews

Hack4Impact partnered with ACT, a child safety non-profit, to create a user-friendly IOS app that digitizes their Safety Deck. This interactive app will replace costly PDFs and physical decks, offering offline access and an intuitive experience for parents, childcare providers, and law enforcement. Over two semesters, we’ll focus on design and development to improve accessibility, reduce costs, and enhance child safety efforts.

Pollen P.I.E

Daniel Baldeo-Thorne, Heisen Kong

Our project develops an autonomous drone for indoor hydroponic farms, addressing pollination and pest/disease detection challenges. The drone navigates confined spaces, automating pollination for flowering crops and using computer vision for real-time pest and disease monitoring. This innovative solution maximizes plant productivity, optimizes growing conditions, and enhances efficiency in indoor agriculture, offering a scalable approach to overcome traditional farming limitations.

CUAir (Cornell University Unmanned Air Systems)

Ashir Rao, Arnav Dangre, Brandon Vasquez Munoz, Justin Xiang, Rhea Goswami

CUAir is an award winning project team that designs and builds autonomous aircraft for real world missions. We are presenting our plane’s Intelligent System, capable of object detection and alphanumeric classification. CUAir is competing in the 2025 C-UASC competition with our new plane, Atlas. The team has 8 subteams and over 60 members in majors such as Computer Science, Electrical and Computer Engineering, and Mechanical Engineering.

Ithaca Community Recovery

Sophie Wang, Afran Ahmed, Joe Ugarte, Tuni Le

Ithaca Community Recovery (ICR) provides care for those facing mental health and addiction challenges. Their current system for managing groups, meetings, and leases is fragmented, making recovery support difficult. In collaboration with Hack4Impact (H4I), ICR is developing an admin-facing platform to streamline administration operations. This platform will not only improve efficiency, but break down stigmas around mental health and make recovery services more accessible.

Nutrivision

Archita Nemalikanti

I want to lose weight, but I need a live assistant to watch what I eat and count my calories because I am too busy to manually add up the numbers using a cell phone app. I built a VR/AR application that watches what you eat, counts your calories, and tracks your health.

CADENA- Coagulant Adsorbent Dosing Network

Johan DeMessie, Ayesha Abdiel

Water scarcity and contamination threaten global health, especially in underserved regions. Traditional water treatment relies on inefficient, costly dosing of coagulants and adsorbents. CADENA integrates an embedded sensor network with discrete complex analysis, CFD modeling, and molecular dynamics simulations to precisely optimize dosing in real time. A multilingual dashboard and mobile app provide live monitoring, making water treatment smarter, cheaper, and more accessible.

Apathia

David Kim, Andy Lasson, Angela Zhu, Lorrin Li, Jeremy Ku-Benjet

Apathia is a tactical online multiplayer mobile game. March your troops across the board and out-maneuver your opponent to claim territory and seize victory!