ECE Master's student @ Carnegie Mellon University · NTU EE Alumnus · Pittsburgh (USA)
- Speech & Audio AI
Spoken LLMs (TASTE), Agentic Audio (Audio-Maestro), Speech-To-Text (Interspeech 2025 Challenge) - Multimodal LLMs
Cross-modal reasoning (DLCV), Large model fine-tuning & LoRA (CVPR 2024 Challenge) - AI × EdTech
AI-assisted learning chat-box (Sophion), Intelligent lecture generation, Entrepreneur, Business development - ML Systems
Cloud-native development (DCmanager-backend), Parallel programming (Parallel-Sokoban-Solver), Scalable inference (MediaTek) - AI × Semiconductor
Wafer detection chat-box (NXP), Digital IC agent (Ultrasystem)
- MediaTek Research — ML Engineer Intern
Co-developed the first speech–text aligned tokenization (TASTE, ICLR 2026) - ASUS AI & Cloud Research — AI Research Intern
Built the Agentic Audio LLM pipeline (Audio-Maestro) and proposed a training-free adaptive steering vector to mitigate audio model hallucination (Interspeech 2026) - NXP Semiconductors — ML Engineer Intern
Built an AI chatbot for wafer defect analysis - EdTech Startup — Founding Engineer
Delivered AI-powered educational solutions (Sophion) · Partnered with top Taiwanese institutions