LLM RAG and Prompt Improvement: Refine Retrieval-Augmented Generation (RAG) and optimize prompt designs for improved interaction outcomes with large language models.Speech Model Interruption Mechanism: Design and develop interruption mechanisms specifically tailored for speech models to ensure seamless handling of real-time interruptions during agent interactions.Kubernetes Deployment: Deploy and manage digital agents in Kubernetes environment to ensure high availability, scalability, and reliability.Container Deployment: Designing and implementing containerized deployment solutions to streamline the deployment process, ensuring consistency and scalability across environments.Coherence improvement and interrupt mechanism design & development for user interactions with digital agents

Human Pose Analysis Project: Developed a real-time pose estimation and human body measurement SDK.Digital Human Project: Developed a real-time digital agent by integrating Text-to-Speech (TTS), Speech-to-Text (STT), Large Language Models (LLM), and Text-to-Image generation models. Designed and synchronized these components to create a seamless, interactive user experience.Sleep Apnea App Project: Created a cross-platform app using Flutter to support couples dealing with sleep apnea. The app raises awareness, encourages consistent usage, and gathers valuable data for research on sleep apnea, providing insights to improve health outcomes.