gpt-bot (RAG)

A Customer support Chatbot built using langchain.
the bot is able to answer user questions about the company products, make comparison between two products, and provide the end user with the pros and cons of each product.
built using dummy hardware csv data and pcandparts website data

how it works

Simply put:

1. We have a collection of data vectorized and stored in a vector database.
2. When a user asks a question, we retrieve the most relevant data from the vector store using similarity search.
3. We pass the most relevant data to the LLM.
4. Finally, the LLM – using relevant data - generates a response.

High Level View

Mid Level View