Module rag 

RAG (Retrieval-Augmented Generation) system with vector stores and embeddings.

RAG (Retrieval-Augmented Generation) Module

This module provides a flexible RAG system with:

• Multiple vector store backends (in-memory, Qdrant)
• Embedding generation (OpenAI API, local models)
• Document chunking and preprocessing
• Semantic search and retrieval
• Reranking capabilities

Structs

Document

Represents a document in the RAG system

InMemoryVectorStore

In-memory vector store using cosine similarity

OpenAIEmbeddings

OpenAI embedding provider using text-embedding-ada-002 or text-embedding-3-small

QdrantVectorStore

Qdrant vector store implementation

RAGSystem

Main RAG system that combines embedding provider and vector store

SearchResult

Represents a search result from the RAG system

Traits

EmbeddingProvider

Trait for embedding generation

VectorStore

Trait for vector storage backends