Custom AI Agent with RAG for Internal Knowledge Base (Python)

We're a mid-size logistics company looking for an experienced AI engineer to build a custom AI agent that can answer employee questions using our internal documents (SOPs, policy manuals, and training guides). The agent should use a RAG (Retrieval-Augmented Generation) setup, pull answers from our document store, and give clear, accurate responses through a simple chat interface. Must be built in Python, and should integrate with our existing Postgres database. Experience with vector databases (like pgvector or Pinecone) and LLM APIs (OpenAI or Claude) is required.