HK Stock IPO Advisory Agent (AI project)
Built an AI agent for HK Stock IPO advisory that orchestrates tool calls to retrieve real-time market data and event calendars. Used MCP-based function/tool integration to support downstream LLM analysis and response generation. Implemented prompt chaining to coordinate retrieval, structured reasoning, and advisory report writing for LLM outputs. • Tool/function orchestration via MCP calls • RAG-like workflow for retrieving up-to-date quotes and reports • Prompt chaining pipeline for structured analysis then natural-language advice • Multi-step advisory generation covering valuation, sentiment, risk, and peer comparison