Chatbot designing

A chatbot can help us in different aspects. Although there are different ways to design a chatbot, the general pipeline is quite similar. For example, if one wants to design an AI chatbot for specific topics, several steps can be embedding, RAG and using LLM API. The embedding part is for more accurate data fetching result. RAG, Retrieval-augmented generation (RAG) is the part for enhancing LLM searching. LLMs can be switched, and the main purpose is to understand and to respond the user. On the other hands, there are already many services online and ready for subscriptions, but building our own framework can also be interesting and helpful. So, what can it be for chatbots? Let’s discuss about this in the future articles.