Unified RAG pipeline for investment intelligence

Modular claim-first retrieval + grounded synthesis powering chat, memo generation, and research pages.

High-level overview of production work (no proprietary details):

  • Unified fragmented LLM calls into a modular RAG pipeline (query optimisation → retrieval → grounded answer synthesis with citations)
  • Multi-provider LLM client for Azure OpenAI + AWS Bedrock
  • Router + evaluation instrumentation for quality monitoring