Developing an Evidenced Counseling Chatbot with LLMs
In the world of AI-driven conversational agents, the issue of “hallucination” remains a significant challenge. This problem becomes particularly acute when dealing with sensitive topics such as health issues and psychology, where misinformation can have severe consequences. In this challenge, we seek to tackle this issue by creating an Evidenced Counseling Chatbot, leveraging Large Language Models (LLMs) and the Semantic Kernel framework from Microsoft, to provide empathetic support while simultaneously offering credible and contextually relevant references from a knowledge database.
Participants in this hackathon will focus on creating a Counseling Chatbot that not only engages in empathetic conversations but also substantiates its responses with reliable references. The key components include minimizing hallucination, integrating knowledge retrieval mechanisms, fostering empathetic communication, and designing a user-friendly interface. By leveraging LLMs and the Semantic Kernel framework, we empower individuals to engage in supportive, informed dialogues about sensitive topics, promoting responsible AI practices and enhancing the overall quality of AI-driven counseling services.