Skip to main content

Introduction to Ragapi

Welcome to Ragapi! Ragapi is a straightforward API that makes it easy for developers to enable AI-powered conversations with documents. With Ragapi, you can process a document (like a PDF, GitHub repository, YouTube video, or website) and interact with it through a conversation endpoint. It’s designed to be fast, simple, and secure. Ragapi only processes data to provide functionality—it does not store any documents or chat history. Instead, all data storage is managed by the user, with documents stored in the user’s own Pinecone database. Additionally, OpenAI does not train models on API data, ensuring your information remains private and secure.

How It Works​

Ragapi operates in two main steps:

  1. Storing the Document: You can call the /store-document endpoint and pass a link to a document. This can be a PDF, a GitHub repository, a YouTube video, or a website URL. Ragapi processes the document, optimizes it for AI interactions, and stores it in a vector database (using the user’s Pinecone database).

  2. Chatting with the Document: Once your document is processed and stored, you can use the /conversation endpoint to ask questions or have a conversation with the document. Ragapi retrieves the relevant information and provides AI-generated responses to your queries.

Key Features​

  • Document Variety: Store different types of content, including PDFs, GitHub repositories, YouTube videos, and websites.
  • Quick Setup: With a single API call, your document is processed and stored, ready for AI interaction.
  • User-Controlled Data: Ragapi does not store any user data, documents, or chat history. Instead, users provide their own OpenAI and Pinecone API keys, keeping data processing and storage fully within the user’s control.

Requirements​

To use Ragapi, you need:

  • A Ragapi API Key: You can generate this key in the Ragapi app.
  • OpenAI API Key: This key is needed for AI processing and answering questions.
  • Pinecone API Key: This key is used to store the document embeddings in your Pinecone database.

With these keys, you’ll be ready to start processing and chatting with your documents in just a few steps.