RAG Collections

Collections organize your documents. Each collection stores documents for a specific purpose.

Creating Collections

Using API

curl -X POST http://localhost/api/v1/rag/collections \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "documentation",
    "description": "Product documentation and user guides",
    "embedding_model": "text-embedding-ada-002"
  }'

Using Python

import requests

response = requests.post(
    "http://localhost/api/v1/rag/collections",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "name": "documentation",
        "description": "Product documentation and user guides",
        "embedding_model": "text-embedding-ada-002"
    }
)

collection = response.json()
print(f"Collection ID: {collection['id']}")
print(f"Collection Name: {collection['name']}")

Using JavaScript

const response = await fetch('http://localhost/api/v1/rag/collections', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    name: 'documentation',
    description: 'Product documentation and user guides',
    embedding_model: 'text-embedding-ada-002'
  })
});

const collection = await response.json();
console.log(`Collection ID: ${collection.id}`);

Collection Parameters

Parameter	Type	Required	Description
name	string	Yes	Unique collection name
description	string	No	Collection description
embedding_model	string	No	Model for embeddings (default: text-embedding-ada-002)

Listing Collections

curl -X GET http://localhost/api/v1/rag/collections \
  -H "Authorization: Bearer YOUR_API_KEY"

response = requests.get(
    "http://localhost/api/v1/rag/collections",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)

collections = response.json()["collections"]
for col in collections:
    print(f"Name: {col['name']}")
    print(f"Documents: {col['document_count']}")
    print(f"Created: {col['created_at']}\n")

Viewing Collection Details

curl -X GET http://localhost/api/v1/rag/collections/documentation \
  -H "Authorization: Bearer YOUR_API_KEY"

response = requests.get(
    "http://localhost/api/v1/rag/collections/documentation",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)

collection = response.json()
print(f"Name: {collection['name']}")
print(f"Description: {collection['description']}")
print(f"Documents: {collection['document_count']}")
print(f"Total Chunks: {collection['chunk_count']}")

Updating Collections

curl -X PUT http://localhost/api/v1/rag/collections/documentation \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "description": "Updated description for the collection"
  }'

response = requests.put(
    "http://localhost/api/v1/rag/collections/documentation",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "description": "Updated description for the collection"
    }
)

print(f"Updated: {response.json()['message']}")

Deleting Collections

curl -X DELETE http://localhost/api/v1/rag/collections/documentation \
  -H "Authorization: Bearer YOUR_API_KEY"

response = requests.delete(
    "http://localhost/api/v1/rag/collections/documentation",
    headers={"Authorization": "Bearer YOUR_API_KEY"}
)

print(f"Deleted: {response.json()['message']}")

Warning: Deleting a collection removes all documents and their embeddings. This action cannot be undone.

Collection Best Practices

Naming Collections

Use clear, descriptive names:

Good Names	Bad Names
product_docs	docs
support_articles	articles
legal_contracts	legal
api_reference	api
research_papers	papers

Organizing Collections

Organize collections by purpose:

documentation/
├── user_guides
├── api_reference
├── troubleshooting
└── faqs

company/
├── policies
├── procedures
└── legal

Collection Settings

Choose embedding model based on use case:

Use Case	Recommended Model
General purpose	text-embedding-ada-002
Code documentation	text-embedding-3-small
High accuracy	text-embedding-3-large

Troubleshooting

Collection Name Exists

Problem: {"error": "Collection name already exists"}

Solution:

• Choose a unique collection name
• Delete existing collection if no longer needed
• Use a more specific name

Collection Not Found

Problem: {"error": "Collection not found"}

Solution:

• Verify collection name is correct
• Check collection exists using list endpoint
• Confirm you have access to the collection

Documents Not Appearing

Problem: Collection shows 0 documents after upload

Solution:

• Wait for document processing to complete
• Check upload response for errors
• Verify file format is supported
• Review server logs for processing issues

Next Steps

• Upload Documents - Add documents to your collections
• Search Documents - Search within collections
• Manage Documents - List and delete documents
• Examples - See collection usage examples