What is RAG? Understanding Knowledge Augmentation Plugins & Their Use with LLMs

 

Did you know if you still do not upgrade your development knowledge, you are no longer considered a developer? Currently, many AI-based development tools are available to handle complex development tasks, but those tools only work best if you have enough knowledge to work with AI.

The retrieval-augmented generation is often confusing for many developers. In reality, it is the type of technology that retrieves relevant information from external sources. For example, you are going to use LLM, a big database of concepts and ideas, but you need to extract a limited scenario that is not general at all. In other words, the LLM is the big brain, and RAG is the reference to the external information.

For example, you have completed your basic schooling, and you know how to use a computer and the Internet. For some reason, you have not joined the university but learned some computer science books to understand some development methods. If you can enroll in a university program and pass out, you will fine-tune your school knowledge. You can deal with any basic math-related tasks.

In case you are unable to take admission, after learning some books, you will gain some knowledge and understand the basic questions. Your limited book knowledge is also helpful in solving various problems. You still have the power to find the answers based on your high school knowledge. It is similar to the RAG, which is the reference to specific knowledge set, and LLM is the big library.

By utilizing RAG, you are basically adding the reference to a specific knowledge set and directing LLM to take action based on RAG reference points.

Why we need to learn the RAG technology, because we want to create a dedicated Agent in the LangChain framework that have personalized use case scenario. Thanks for reading.