The Immunization Information Systems (IIS) program at the CDC has developed many guidance documents, resulting in a vast knowledge base. To better search, summarize, and help users update this information, we partnered with two teams at CDC’s Informatics and Data Analytics Branch to create AI Retrieval Augmented Generation (RAG) proof-of-concept chatbots:

• The CDC HL7 team, who focus on data exchange standards, are full partners design, testing, and quality assurance.
• The CDC CDSi (Clinical Decision Support for Immunizations) team is at the beginning of their journey; we expect their trajectory to mirror the HL7 team’s.

We will discuss the process of working alongside these business teams to create a fit-for-purpose AI Chatbot.

The new chatbots use publicly available, trusted sources to ensure accurate responses, highlighting the importance of high-quality documentation, clear terminology, and structure. These proof-of-concept bots are currently improving the efficiency of managing guidance documentation.

LEARNING OBJECTIVES

• Discuss how AI can assist and accelerate updating high-accuracy content.  
• Learn how to lead teams from nascent AI understanding to being advocates for AI
• Understand the importance of good education and culture in rolling out AI initiatives