RAGify Documentation

RAGify is an AI-powered document analysis framework that combines the capabilities of large language models with efficient document retrieval. It enables intelligent document processing, semantic search, and contextual question answering.

Fast Processing

Process large documents in seconds with optimized chunking and parallel processing capabilities.

Smart Analysis

Leverage GPT-4 for deep document understanding and intelligent question answering.

Vector Search

Utilize Pinecone for efficient semantic search across your document collection.

Getting Started

Follow these steps to set up RAGify in your environment:

Install Dependencies

pip install ragify
pip install openai pinecone-client langchain

Environment Setup

export OPENAI_API_KEY="your-api-key"
export PINECONE_API_KEY="your-pinecone-key"
export PINECONE_ENV="your-environment"

Pinecone Configuration

Configure Pinecone for vector storage and retrieval:

Initialize Pinecone

from ragify import PineconeManager

# Initialize Pinecone connection
pm = PineconeManager(
    api_key="your-api-key",
    environment="your-environment"
)

# Create vector index
pm.create_index("docs-index", dimension=1536)

# Connect to existing index
index = pm.connect_index("docs-index")

OpenAI Integration

Set up document processing with OpenAI embeddings:

Document Processing

from ragify import DocumentProcessor

processor = DocumentProcessor()

# Process single document
doc_embeddings = processor.process_file("document.pdf")

# Process directory
collection = processor.process_directory("./docs")

# Query documents
results = collection.query(
    "What are the key findings?",
    model="gpt-4-turbo-preview"
)

API Reference

Endpoint	Method	Description
/documents/process	POST	Process and embed documents
/documents/query	POST	Query processed documents
/collections	GET	List document collections
/collections/{id}	GET	Get collection details

Example API Usage

import requests

# Process document
response = requests.post(
    "https://api.ragify.com/documents/process",
    files={"file": open("document.pdf", "rb")}
)

# Query documents
response = requests.post(
    "https://api.ragify.com/documents/query",
    json={
        "query": "What are the main conclusions?",
        "collection_id": "col_123"
    }
)

RAGify Docs