From the course: Oracle Cloud Infrastructure Generative AI Professional Cert Prep

Unlock this course with a free trial

Join today to access over 24,900 courses taught by industry experts.

Chatbot technical architecture

Chatbot technical architecture - Oracle Cloud Infrastructure Tutorial

From the course: Oracle Cloud Infrastructure Generative AI Professional Cert Prep

Start my 1-month free trial Buy for my team

Chatbot technical architecture

“

(gentle music) - [Hemant] So far, we have seen how retrieval augmented generation works through lessons and demos. Let us recap it through right now. RAG is split into two parts, indexing and retrieval and generation. We index documents by first loading them. Next we split documents into manageable chunks. Next, we embed the chunks and persist the embedded chunks in the vector databases, like Chroma for FAISS for retrieval later. LangChain offers a variety of document loaders, splitters, and embedding models, and we have used some of these in our demos. We also have seen how we persist embedded chunks into the Vector databases like Chroma and FAISS. For document retrieval, we begin with establishing a connection to a persistent database or loading a database. Next, we create a retrieval chain. Chain uses a retriever and memory to pull relevant documents and chat history by using the retrieval memory. These are sent to the LLM along with a question that is input from a UI like…

Contents