RAG Engine

RAG Engine makes building AI apps that need external data a breeze. With just a few simple API calls, managing data is easy peasy. One call adds a data source, and another searches all your data. This makes it great for both tech-savvy folks and those who aren''t.

Key Features

RAG Engine shines with its simple data management. Add a data source with an API call that includes your project API key and namespace. This keeps user data separate. Here''s how you add a data source:

curl -X POST https://ragengine.io/data-sources 
-H "Content-Type: application/json" 
-H "Authorization: Bearer YOUR_PROJECT_API_KEY" 
-d '{ "namespace": "user:1", "type": "website", "url": "https://mywebsite.com" }'

To search your data, make another API call:

curl -X GET "https://ragengine.io/search?query=...&namespace=user:1" 
-H "Authorization: Bearer YOUR_PROJECT_API_KEY"

The response shows documents with their sources, content, and similarity scores:

{
"documents": [
{
"source": {
"type": "website",
"url": "https://mywebsite.com"
},
"content": "...",
"similarity_score": 3.99
},
{
"source": {
"type": "file",
"filename": "pricing.txt"
},
"content": "...",
"similarity_score": 3.11
}
]
}

RAG Engine handles all the background work. It manages data ingestion, processing, storage, and synchronization. You don''t need to worry about any infrastructure. Everything is hosted by RAG Engine.

Benefits

RAG Engine is designed to be affordable. It scales with your usage, moving to higher-tier instances as needed. The pricing is straightforward:

Cost/Price

1. Vector Database: Starts at $4/month for a DigitalOcean droplet (512 MB RAM, 10 GB SSD). Larger datasets need more resources, but it''s always at-cost with no markup.
2. Embeddings: Exact OpenAI rates are passed along ($0.010 to $0.065 per 1M tokens), with no extra fees.
3. RAG Engine Fee: $4.99/month (first month free). This covers the ingestion pipeline and unified search.

Use Cases

RAG Engine supports website URLs and various file formats. Plans include adding integrations for Notion, Google Docs, and other platforms based on user feedback.

Vertex AI''s RAG Engine is a fully managed service. It helps build and deploy RAG implementations with your data and methods. Key features include:

• Flexible Architecture: Choose models, vector databases, and data sources that fit your needs.
• Evolving Use Cases: Add new data sources, update models, or adjust retrieval parameters easily.
• Evaluation: Set up multiple RAG engines to find the best configuration for your use case.

Customization

• Parsing: Adjust chunk size and overlap for different document types.
• Retrieval: Works with vector storage options like Pinecone, Weaviate, or Vertex AI Vector Search.
• Generation: Choose from many LLMs in Vertex AI Model Garden, including Google''s Gemini, Llama, and Claude.

Integration with Gemini

Vertex AI''s RAG Engine integrates with the Gemini API. Create grounded conversations that use RAG for relevant answers. Initialize a RAG retrieval tool with specific settings and pass it to a Gemini model.

RAG Engine makes building AI applications easier. It offers a managed service with flexible and customizable options. This is great for small startups and indie developers. For more information or to join the waitlist, visit RAG Engine or check their Discord server for support.