AI Agent Intelligent Customer Service Project Labeling Contributor
Text
Question AnsweringParticipated in the development of an intelligent customer service AI agent system using prompt engineering and knowledge-base Q&A methods. Labeled and structured Q&A pairs for e-commerce customer service dialogues, enabling the model to accurately respond to logistics, refund, order status, and after-sales queries. Designed labeled prompts and responses to improve the agent's function calling and knowledge retrieval abilities. • Created and evaluated prompt-response pairs specifically for e-commerce customer inquiries. • Annotated scenarios involving disputes, complaints, and refund requests for AI model training. • Organized and labeled text data based on real store knowledge bases and frequent user questions. • Assisted with the deployment and testing of labeled data within the Streamlit demo system.