2024 San Diego Fall Classic: 3rd Place in Men’s Open Novice 4+

• Assisted research in applying Large Language Model in multi-issue negotiations, leveraging contextual reasoning while addressing arithmetic limitations through linear programming, enhancing strategic decision-making in bargaining scenarios• Implemented priority consistency checker using a chain-of-thought prompting scheme and OpenAI API to ensure alignment between LLM agents’ offers, stated priorities, and negotiation history

• Developed user-centric features and backend systems for a beauty retail app, optimizing code scalability and cleanliness, and contributed to the software development lifecycle. • Collaborated with cross-functional teams, employing Agile practices to effectively deliver software under tight deadlines, and engaged partners to design solutions based on technical requirements and business use cases.• Designed a skin analysis function using Flutter, enhancing feature precision with real-time face detection and advanced camera controls, and improved analysis accuracy by integrating iOS and Android native codes.• Enhanced machine learning model performance using TensorFlow, optimized data sets to reduce error rates, and maintained software efficiency by implementing updates and employing real-time monitoring tools.• Managed application deployment and testing processes, achieving 86% code coverage and reducing deployment time using automated tools, and ensured continuous system efficiency with periodic optimizations.

• Built and optimized a secure, scalable Android-based access control application for construction sites, enhancing efficiency and reducing costs through digital transformation. • Implemented on-device facial recognition using TensorFlow Lite and SIMD operations in C++, improving latency and processing speed for real-time access control.• Deployed AI models on Android devices, ensured robust performance under diverse conditions, and maintained high security and privacy standards by encrypting sensitive biometric data.

• Developed an Android-based access control app for a construction site, utilizing facial detection AI to streamline access management, boosting efficiency by 27%• Collaborated with AI researchers to enhance facial recognition reliability, reducing customer complaints by 32%• Implemented admission thresholds and model improvements to enhance facial recognition robustness