Build an Application to Chat with Documents Using RAG
Learn how to use PostgreSQL pgvector and langchaingo to implement RAG.
We'll cover the following...
Let’s begin by applying the RAG technique to enhance the chat application further.
First, the contents of the chapter from ‘War and Peace’ are converted into vector embeddings and loaded into a PostgreSQL database.
When a user interacts with the chat application and asks questions about the chapter, RAG is used behind the scenes to get a response and send it back to the user.
Set up managed PostgreSQL service
Note: Please skip this section if you already configured the PostgreSQL service from the previous lessons.
To set up the PostgreSQL service:
Sign up for a free Aiven account and log in to Aiven console.
Create a PostgreSQL service.
In the Aiven console, go to the organization > project > Aiven for PostgreSQL service. On the “Overview” page of the service, under “Connection information,” copy the “Service URI.”
Note down the Service URI because it will be used in subsequent steps.
Set up OpenAI API
The OpenAI embedding model (text-embedding-3-small) will be used to create vector embedding for the contents of the chapter from ‘War and Peace.’
Note: If you already have an OpenAI account and API key, please skip this section. ...