RAG Module¶

Overview¶

The RAG (Retrieval Augmented Generation) module enhances AI responses by retrieving relevant information from documents to provide context-aware answers. It implements a retrieval-augmented generation pipeline.

Module Structure¶

rag/
├── rag.module.ts       # Module definition and configuration
├── rag.service.ts      # Core RAG implementation
└── dto/                # Data transfer objects
    └── rag-service.dto.ts # DTOs for RAG service operations

Key Components¶

RagModule¶

The root module that configures RAG services.

Imports: None
Providers: RagService
Exports: RagService

RagService¶

Implements the RAG pipeline: - Question reformulation for improved retrieval - Document retrieval from vector store - Context-aware answer generation - Integration with LangChain and Pinecone services

DTOs¶

Data transfer objects for RAG service operations: - QuestionResponseSchema: Defines the structure for question responses

Dependencies¶

The RAG module depends on: - LangchainModule: For LLM interaction - VectorstoreModule: For document retrieval from vector database