Project Experience

AI Applications for Freshmen Assistance (Based on Large Language Models)

The University of Sheffield 02/2024~06/2024

Participated in developing an AI chatbot on IBM Watson, creating and optimizing datasets for accurate responses to freshman inquiries, enhancing their university experience.
Integrated the ChatGPT Assistant API with IBM Watson, broadening the AI's knowledge base and improving its capability to address complex student issues effectively.
Ensured the AI assistant provided a seamless and efficient interaction, with a focus on user experience and precision in answering a wide range of university-related questions.

The University of Sheffield 02/2024~05/2024

Developed navigation algorithms on Waffle Pi, using Rospy, OpenCV, and Numpy library for real-time decision-making, allowing robots to navigate with enhanced precision and efficiency autonomously
Implemented a computer vision system for dynamic target detection and SLAM technology for real-time mapping, reinforcing robots' ability to adapt and respond

The University of Sheffield 02/2024

Developed "Vari-Sign," a system that uses real-time data to adjust speed limits, improving rural road safety through a cross-disciplinary approach
Utilized advanced sensors and Orange Pi microprocessors for real-time weather and road data collection, crucial for dynamic speed limit adjustments, reducing traffic accidents
Developed "Vari-Sign," a system that uses real-time data to adjust speed limits, improving rural road safety through a cross-disciplinary approach

The University of Sheffield 10/2023~12/2023

Collaborated in a team of four to develop a software system for "Trains of Sheffield," a retailer of model railways, focused on managing inventory, sales, and customer interactions.
Analyzed business requirements, designed a UML class diagram, and developed a normalized data model to capture complex relationships between products, orders, and user roles.
Implemented a MySQL database, Java middleware for business operations, and a Java Swing user interface to enable smooth interactions for customers, staff, and managers.
Conducted thorough security assessments, including SQL injection prevention, secure password storage, and role-based access control to ensure system robustness.

Github: Train-model-Shop

The University of Sheffield 10/2023~12/2023

Developed a solution to classify chessboard diagrams from printed images, identifying the location and type of each chess piece on the board.
Employed feature extraction and dimensionality reduction techniques to preprocess images, reducing each square to 10 key features while handling noisy and clear datasets.
Achieved robust performance on both clean and noisy data, enhancing the classifier's ability to handle variability in image quality due to historical printing imperfections.

The University of Sheffield 2/2023~5/2023

Collaborated in a team to develop EdFlix, a course discovery platform aimed at helping users find and manage online training courses tailored to their interests and backgrounds.
Designed and implemented key features, including a course suggestion system, user account management, and an AI-based recommendation feature to enhance user engagement.
Worked on course management modules, enabling users to rate, review, and suggest courses, while allowing moderators to tag, update, and approve new courses based on user requests.

The University of Sheffield 1/2023

Collaborated to design an eco-friendly “Eco Planter," integrating hydroelectric and UV LED technologies for social housing in Glasgow, enhancing energy efficiency and food security for low-income residents
Played a crucial role in developing a system that sustainably meets the energy and nutritional needs of communities, reducing environmental impact and dependence on costly imports
Conducted STEEPLE analysis and engaged with the local community to refine the project, ensuring it was culturally responsive and economically viable for Govan’s unique context