AI Agent and Knowledge Base Developer (Multi-Agent Decision Support System)
Text
Text GenerationIndependently designed and developed a multi-agent decision support system focused on RAG-based private domain research. Responsible for structuring uploaded user data, hybrid search, semantic retrieval, and context packaging for downstream AI analysis. Developed, tested, and iteratively improved the knowledge base and output mechanisms for enhanced reliability and trustworthiness. • Designed, labeled, and structured user-contributed text and research documents into a persistent knowledge database for multi-context AI research tasks. • Labeled relevant sources, quotes, and responses for secondary verification and context integrity, including reference hit rate enhancement through structured annotation. • Built workflows for multi-role agent discussion and adversarial role-play, annotating roles, responses, and decision outputs for AI evaluation and training. • Performed fine-grained text labeling with retrieval cues, structured summaries, and citation mapping to support prompt engineering and structured outputs.