1. Project Overview

SparkX is developed using Springboot3 + PgSQL + Vue3, and uses large language models and RAG for knowledge base question-answering systems. It is plug-and-play, model-agnostic, and flexible, supporting rapid integration into third-party business systems.

2. Open Source License

Licensed under GPL-3.0

3. Screenshots

UI Overview

4. Feature Overview

SparkX is developed using Springboot3, a knowledge base question-answering system based on large language models and RAG. It is plug-and-play, model-agnostic, and flexible, supporting rapid integration into third-party business systems.

Software Features

• Model-agnostic
  Using the langchain4j framework, it is flexible to connect to common frameworks on the market. The interface is unified, making it easy to extend. It also introduces the OpenAI standard interface, which can connect to almost any mainstream large model, not limited to official services, and can also support self-built models.
• RAG index enhancement technology
  Using RAG index enhancement technology, it effectively eliminates the hallucinations of large models, and enhances the knowledge accuracy of large models, especially effective for private data question-answering.
• Flexible orchestration
  Using AntvX6 visualization workflow orchestration engine, it is flexible to extend the practicality of applications. It allows multiple agents to work together, making it easier to complete user consultations.
• Easy embedding
  The system supports two deployment methods: single-page and pop-up embedding methods. It can instantly improve the service capabilities of the embedded system, allowing the embedded system to be easily integrated into AI without any modification.
• Unlimited plugin extension
  : System supports customizable plugins, allowing for flexible interaction with AI through self-owned API, thereby providing accuracy and professionalism in replies. Additionally, through the powerful MCP, AI capabilities can be extended without restriction.

5. Tech Stack

Technical Architecture

Language: Java17 + JavaScript + Node.js
Framework Tools: Springboot3 + Langchain4j + Vue.js + ElementUI PLUS + NPM + Maven
Database: PgSQL (vector extension) + MapDB
Development Tools: IntelliJ IDEA + WebStorm + Navicat

Local Development Suggested Versions
Java17, Node.js v22.17.0, NPM 10.9.2, PgSQL 15, Navicat Premium Lite 17
Unmentioned, can use any version or the version already agreed upon in the project.

Quick Start

1. Download source code

git clone spark-x

2. Enter the docker directory and install

cd spark-x/docker
docker compose up -d

3. Access

http://localhost:8189

Account admin
Password admin