This project is ongoing🔥🔥🔥

Open BI Platform

An open-source business intelligence platform based on microservice architecture.

Introduction · Architecture · AI Analysis Workflow · RabbitMQ Workflow · Open AI · Tech Stack · Setting Up Locally · Project Overview · License

Check Frontend Source Code 👀

Check Python Poe Service Source Code 👀

Introduction

An open-source business intelligence platform based on microservice architecture. By leveraging ChatGPT technology, users only need to import raw Excel data and enter their analysis requirements. The system will then automatically generate visual charts and draw analytical conclusions, significantly reducing the cost of manual data analysis.

Architecture

AI Analysis Workflow

RabbitMQ Workflow

Open AI

This project uses the ChatGPT service provided by Poe. I created an Open BI Platform Bot based on ChatGPT 3.5 in Poe. The user's input will be processed by the backend into custom prompt requirements for the Open BI Platform Bot, and the returned results include Apache ECharts chart code and data conclusion. Incredibly, with the support of custom prompts, the Open BI Platform Bot can generate result formats with up to 100% accuracy🔥. This means the chart code can be directly used by the front end, solving the problem that different data requires different charts and backends cannot handle the pain points of all cases.

Besides, thanks to snowby666 for the open source Poe reverse engineering tool poe-api-wrapper. Based on this tool, I built a RESTful API for calling the Open BI Platform Bot. The API is written using Python and the Flask framework and handles multiple requests in a multi-threaded manner. If you are interested, you can check out the open-bi-platform-poe-service project.

Open BI Platform Bot Preview:

Code snippet calling the Open BI Platform Bot service:

package com.glenncai.openbiplatform.aianalytics.manager;

import cn.hutool.core.lang.TypeReference;
import cn.hutool.json.JSONUtil;
import com.glenncai.openbiplatform.aianalytics.exception.enums.AiExceptionEnum;
import com.glenncai.openbiplatform.aianalytics.model.dto.ChatRequest;
import com.glenncai.openbiplatform.aianalytics.model.dto.ChatResponse;
import com.glenncai.openbiplatform.aianalytics.utils.HttpUtils;
import com.glenncai.openbiplatform.common.common.BaseResponse;
import com.glenncai.openbiplatform.common.constant.AiConstant;
import com.glenncai.openbiplatform.common.exception.BusinessException;
import lombok.extern.slf4j.Slf4j;
import org.springframework.stereotype.Service;

/**
 * Communicate with AI service (e.g. OpenAI, Google Bard, POE, etc.)
 *
 * @author Glenn Cai
 * @version 1.0 07/26/2023
 */
@Service
@Slf4j
public class AiManager {

  /**
   * Process AI chat
   *
   * @param chatRequest chat request body, including 'message' and 'key'
   * @return AI response
   */
  public String doAiChat(ChatRequest chatRequest) {
    String jsonStr = JSONUtil.toJsonStr(chatRequest);
    String result = HttpUtils.post(AiConstant.AI_API_URL, jsonStr);
    TypeReference<BaseResponse<ChatResponse>> typeRef = new TypeReference<>() {
    };

    BaseResponse<ChatResponse> response = JSONUtil.toBean(result, typeRef, false);
    if (response == null) {
      throw new BusinessException(AiExceptionEnum.AI_RESPONSE_ERROR.getCode(),
                                  AiExceptionEnum.AI_RESPONSE_ERROR.getMessage());
    }
    if (response.getCode() != 0) {
      throw new BusinessException(AiExceptionEnum.AI_RESPONSE_ERROR.getCode(),
                                  AiExceptionEnum.AI_RESPONSE_ERROR.getMessage());
    }

    return response.getData().getContent();
  }
}

In the future, I will integrate this AiManager into an SDK to facilitate the use of other projects.

Example of results returned by Open BI Platform Bot:

Tech Stack

Backend

• Java
• Spring Boot
• Spring Cloud Gateway
• Spring Cloud Config
• Spring Cloud Circuit Breaker
• Spring Cloud Load Balancer
• Spring Cloud Sleuth
• Spring AOP
• Spring Cache
• Eureka
• OpenFeign
• RabbitMQ
• Redis
• MySQL
• MyBatis-Plus
• Zipkin
• Docker

Frontend

• Next.js
• React
• TypeScript
• Tailwind CSS
• Apache ECharts

Python AI service

• Python
• Flask
• poe-api-wrapper

Setting Up Locally

For Microservices Backend:

1. Java 17
2. MySQL 8.0+
3. Redis
4. RabbitMQ
5. Zipkin

For Frontend:

1. Node.js 18.0+
2. npm 9.0+

For Python AI service:

1. Python 3.9

docker-compose -f docker-compose.env.yml up -d

docker-compose -f docker-compose.service.yml up -d

Project Overview

Many APIs have been implemented on the backend, while the frontend still does not call these APIs. The project is still in progress.

Authentication:

Dashboard:

AI Analytics:

User Management:

My Account:

Feedback:

Showing Responsive Design:

License

Licensed under the Apache-2.0 license.