Augmenting a LangGraph Chatbot with a Dictionary Service

Teaching the chatbot medical acronyms and synonyms

Drug trials are research studies conducted to evaluate the safety, efficacy, and effectiveness of drugs. Notably, they offer patients access to potentially life-changing treatments before they’re widely available. However, many individuals remain unaware of these opportunities due to the complex nature of trials and the lack of easily accessible information. By leveraging the power of chatbot technology and a robust drug trial database, we can bridge this knowledge gap and empower patients to make informed decisions about their healthcare options.

To address this unmet need, I developed DrugDB (1), a DuckDB-based drug database, and DrugBot (1), a LangGraph chatbot interface for it. DrugDB is a multi-modal database, supporting SQL, graph, vector, and full-text search capabilities. DrugBot simplifies user interaction by providing an intuitive web interface (Figure 1) that allows users to query DrugDB using natural language.

Figure 1. A dialog in the LangGraph DrugBot. Image by author.

However, because DrugBot directly interfaces with the database, it requires users to input precise medical terminology. As…