Hello there, I'm so happy you are here! (:My name is Yuliana, I'm a master computer science student based in Texas. I have a strong passion for the technologies related to SWE and AI.Outside of my regular activities, I'm a sneaker enthusiast, coca-cola aficionado, ping pong player, and a guinea pig lover.I'm continuously thrilled to partner with individuals and teams who wholeheartedly embrace my fervor for crafting extraordinary experiences. Your visit to my portfolio is deeply appreciated !You can download my resume here.
AI Tutoring System, developed during Hack Research 2023 and took the 1st place win in AI Healthcare, is an innovative app that uses GPT-4 and advanced data extraction techniques to enhance healthcare education. By combining web scraping, embeddings, and retrieval-augmented generation, it sets a new standard for AI-driven learning tools in the medical field.
In Bin Fiesta, I developed a Google Gemini AI-powered chatbot that provides users with personalized recycling guidance, contributing to our 1st place win in Sustainability at Frontera Hacks. The project combines real-time AI interaction with a user-friendly interface built in Next.js, encouraging eco-friendly practices through accessible recycling information.
The project, Balanced Perspective, aims to use deep learning for unbiased news summarization. By leveraging a CNN classifier to detect political bias and the PEGASUS transformer model for abstractive summarization, it produces summaries intended to present balanced perspectives across articles from different political spectrums​.
This project focuses on transferring data from an original SQLite database to a new relational schema. It involves creating tables for musicians, albums, and instruments, importing data from a CSV, and setting up relationships between these tables to support efficient data retrieval and reporting.
This project utilizes Fourier transforms to analyze and process a lung CT scan, extracting magnitude and phase spectrums and downsampling the image. Linear interpolation and zero-padding techniques are then applied to upsample the image, with mean squared error calculated to evaluate the quality of the interpolated images.
I developed a book club management app providing the organizer, with full CRUD functionality to manage book suggestions, track reading status, and approve or reject recommendations from members. The app allows members to sign up, suggest books, and view read/unread lists, while the organizer can control the visibility of suggestions to maintain a high-quality selection process.
I built a Discord bot that responds directly when users type in, #chat, followed by their message. This bot enhances interaction by providing instant replies within the same message thread, making communication seamless and engaging.