Skip to content

LaoM10617/AI-System-Architechture-Designer

 
 

Repository files navigation

A lightweight AI-Powered System Architecture Design Assistant

A lightweight, interactive tool that helps users design software system architectures through AI-powered suggestions, visual diagrams, and notes editing.

Modern Dashboard

Overview

This application provides a whiteboard-style interface where users can:

  • Interpret project visions in detail through interactive sessions
  • Identify key modules and refine requirements with AI guidance
  • Automatically generate structural diagrams and documentation
  • Lower communication and decision-making costs in architecture design

Key Features

Control Panel

  • API Choice: Support for both OpenAI and Gemini AI backends
  • Feature Customization:
    • Application type selection
    • Core feature selection
    • Project description input
  • Generation Types:
    • Text output in sticky notes
    • Architecture proposals
    • Visual diagrams

Whiteboard Interface

  • Sticky Notes: Add, delete, edit, and freely move notes
  • Bins System:
    • Trash Bin: Delete, permanently remove, or restore notes
    • Doc Bin (Favorites): Store important information
    • Visibility control for both bins

AI-Powered Capabilities

  • Architecture suggestion generation
  • Mermaid.js diagram creation from architectural descriptions
  • Multiple-choice question generation to clarify project requirements
  • Note content suggestions based on existing context

Recent Updates

MCQ Generation

Diagram Generation and Editing

Installation & Setup

Backend Setup

  1. Install Python dependencies:
pip install -r requirements.txt
  1. Create an api.env file with your API key:
OPENAI_API_KEY=your_openai_api_key
# OR
GEMINI_API_KEY=your_gemini_api_key
  1. Start the backend server:
python backend.py # For using OpenAI
# OR
python backend_Gemini.py # For Using Gemini
  1. Open index.html in the browser

About

PIIS25_PraktikumProjekt

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Jupyter Notebook 44.0%
  • HTML 28.3%
  • JavaScript 21.1%
  • CSS 4.4%
  • Python 2.2%