Collaborative augmented language models for capturing and utilizing domain expertise

This invention describes a system that creates specialized AI assistants, called digital subject matter experts (dSMEs), by capturing knowledge from human experts through conversations and then organizing it. When a user asks a question in natural language, a chatbot routes the request to the most relevant dSME by comparing the request to descriptions of the available dSMEs using vector similarity. The selected dSME then attempts to answer the question or guide the user, potentially using external tools if its own knowledge is insufficient.