Retrieval-Augmented Generation (RAG) is an AI framework that combines large language models (LLMs) with external knowledge bases to enhance the accuracy, relevance, and reliability of generated text.

LangChain is a framework designed to facilitate the development of applications that leverage large language models (LLMs) through RAG. It provides a structured approach to building RAG applications, which includes:

• Indexing: This involves loading data, splitting it into manageable chunks, and storing it in a vector database for efficient retrieval;
• Retrieval and Generation: LangChain employs a retrieval mechanism to fetch relevant information based on user queries;
• Integration with External Data: By connecting to various data sources, LangChain enables LLMs to access real-time information beyond their training data;

Gemini is a family of multimodal large language models (LLMs) developed by Google DeepMind, designed to process and understand various types of data, including text, images, audio, and video.

Source: https://codelabs.developers.google.com/static/multimodal-rag-gemini

In our guide we used ChromaDB as a vector database, LangChain as a framework for setting up the RAG system integrated with the Gemini Pro / Flash LLM. Source code and full-guideline can be found in the notebook below.

Kaggle Notebook: https://www.kaggle.com/code/armanzhalgasbayev/rag-llm-langchain-gemini-ai

• Karndeep Singh. "Langchain_With_Gemini_And_Build_RAG.ipynb";
• https://www.langchain.com/
• https://aistudio.google.com/
• https://www.perplexity.ai/