Customer Support RAG Service

Overview

Please note, this project is a prototype and not intended for production use.

• A Spring Boot application providing intelligent customer support using Retrieval-Augmented Generation (RAG)
• Uses vector store for semantic document search

Key Features

• Document Upload

  • Supports markdown and text files
  • Content-based duplicate detection using SHA-256 hashing

• Vector Storage

  • Semantic search capabilities
  • Efficient document retrieval
  • Metadata tracking for each document chunk

• REST API

  • Document management endpoints
  • Query interface for support questions
  • Swagger/OpenAPI documentation

Prerequisites

• Java 21 or higher
• Maven 3.6+
• Vector database (Postgres with pgvector)
• Ollame with mistral model for chat and nomic-embed-text for embeddings

Quick Start

Installation Steps

1. Clone the repository to your local machine
2. Run mvn clean install to build the project
3. Start the application using mvn spring-boot:run
4. Access the application at http://localhost:8080

All dependencies will be automatically downloaded and started during the build process. You can see dependencies in docker-compose.yml file

Swagger UI

• Available at: http://localhost:8080/swagger-ui.html
• Interactive API testing
• Complete endpoint documentation
• Request/response examples