UC Santa Barbara Computer Science Capstone Presentations March 14, 2025 in Henley Hall
Sponsor: Acelot
Team: Vincent Zhu, Bryan Huang, Kevin Deng, James Fu, George Nebieridze
Mentor(s):
Lewis Whitehead, Kevin Burk
Project Overview: In our capstone project we will be working on a drug discovery algorithm that
calculates the location and velocity of atoms. The current code base runs a single thread of a CPU and our
job is to
optimize the code by creating a thread secure algorithm that will run on a 9000 core GPU.
Vision Statement:
[PDF]
Project Requirements Document:
[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster:
[PDF]
Final Presentation:
[PDF]
Sponsor: Appfolio
Team: Kole Kikuta, Sid Rambhatla, Koray Kondakci, Emre Cikisir, Andrew Thai
Mentor(s):
Wade Varesio, Jacqui Mai
Project Overview: We are building an AI-powered tool to make property management easier by automating
tasks like tenant questions, lease management, and maintenance requests. We are working to help property
managers
save time and streamline communication with their tenants, making day-to-day operations smoother and more
efficient.
Vision Statement:
[PDF]
Project Requirements Document:
[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster: [PDF]
Final Presentation: [PDF]
Sponsor: Appfolio
Team: Hugo Lin, Liliana Nguyen, Winbert Zhang, Riley Chou, Krish Chaudhary
Mentor(s):
Wade Varesio, Jacqui Mai
Project Overview: We are building an anomaly detection model to detect transaction anomalies in areas
like the history, location, and payment amount. We will deploy the model as a service for AppFolio along
with a web
application as a user interface to view anomalous transactions.
Vision Statement:
[PDF]
Project Requirements Document:
[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster:
[PDF]
Final Presentation:
[PDF]
Sponsor: Artera
Team: Kaitlyn Tom, Michael Wong, Zhenyu Yu, Yuji Sakaguchi, Connor Duncan
Mentor(s):
Anav Sanghvi
Project Overview: ConnectAble is an AI-powered VR social interaction simulator designed for
therapists to assess and support patients with social anxiety. We analyze eye contact, body language, and
speech patterns in
real time to provide objective feedback and measurable progress tracking. This allows therapists to deliver
more effective, data-driven treatment, helping patients build confidence and improve their social skills in
a
controlled, immersive environment.
Vision Statement:
[PDF]
Project Requirements Document:
[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster:
[PDF]
Final Presentation:
[PDF]
Sponsor: Ava Labs
Team: Adyah Rastogi, Wesley Truong, Justin Lang, Hao Lee, Jiahua Ren
Mentor(s):
Aaron Buchwald
Project Overview: Our project seeks to enhance the underlying architecture of the Avalanche
blockchain by optimizing how key-value pairs of Merkle trees are stored directly on disk, moving beyond
conventional database
solutions like LMDB, LevelDB, and RocksDB. By developing and refining critical components such as the disk
manager, free list, and revision manager, we aim to streamline the storage of recent Merkleized blockchain
states,
significantly reducing overhead and improving efficiency in state management.
Vision Statement:
[PDF]
Project Requirements Document:[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster: [PDF]
Final Presentation: [PDF]
Sponsor: Beegol
Team: Edwin Yee, Ray Du, Baimin Wang, Zixiao Jin, Alexander Nguyen
Mentor(s):
Gilberto Mayor
Project Overview: The goal of our project is to create software to provide network diagnostics for
connection drops and disconnects. Using Large Language Models (LLMs), we seek to translate error logs into
human-readable messages and formulate a hypothesis for the root cause of the network problem.
Vision Statement:
[PDF]
Project Requirements Document:[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster:
[PDF]
Final Presentation:
[PDF]
Sponsor: FOIA-Friend
Team: Shelly Zhu, Winston Zuo, Erica Liu, Christian Baker, Ritu Kirsur
Mentor(s):
Nicolaas Jedema, Irvin McCullough
Project Overview: The FOIA Rater and Editor (FARE) project simplifies the Freedom of Information Act
(FOIA) request process by providing users with tools to evaluate and improve their submissions. Through a
rater that
assesses the quality and clarity of FOIA requests and an editor that offers improvement suggestions, FARE
enhances request accuracy, reducing delays and costs for both users and government agencies.
Vision Statement:
[PDF]
Project Requirements Document: [PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster: [PDF]
Final Presentation: [PDF]
Sponsor: Mysten Labs
Team: Sophia Weiler, Ashton Wong, Chloe Ta, Ben Richardson, Bonnie Huynh
Mentor(s):
Deepak Maram, Alberto Sonnino
Project Overview: SuiChat is a messaging app prototype that circumvents the need for intermediaries
in contact discovery and message transmission. Utilizing blockchain technology, the app will facilitate the
establishment of end-to-end encrypted sessions and message transmission, mitigating the censorship and
coercion risks present in current messaging platforms.
Vision Statement:
[PDF]
Project Requirements Document:[PDF]
Project Requirements Document v2:
[PDF]
End of CS189A (First Quarter) Presentation:
[Slides]
[Video]
Development Plan:
[PDF]
Final Poster:
[PDF]
Final Presentation:
[PDF]