generative-ai video for a Generative AI Engineer is optimizing a customer support chatbot that uses a RAG architecture. Users frequently ask ambiguous questions
A Generative AI Engineer is optimizing a customer support chatbot that uses a RAG architecture. Users frequently ask ambiguous questions like “Why was my order late?” or “How do I reset it?” which lack sufficient context. The engineer wants the chatbot to automatically enrich these queries with relevant user-specific information (e.g., order ID, product type) stored in a structured database. What is the BEST strategy to implement this enrichment while preserving model accuracy and performance?