BOOT is an experimental full-stack AI application built to explore how large language models can be integrated into modern web applications and used to power intelligent application workflows.
The project combines a React-based frontend with an Express backend and a dedicated LLM orchestration service, providing a foundation for experimenting with model-driven interactions, AI-assisted features, and intelligent web experiences.
React + Vite Client
│
│ API Requests
▼
Express Server
│
▼
LLM Orchestrator
│
▼
Language Model / AI Services
The separation between the frontend, backend, and LLM orchestration layer is intended to make the AI components easier to evolve independently from the user interface.
- AI-Powered Workflows — Designed to integrate language-model capabilities into application logic.
- LLM Orchestration — Dedicated backend service for coordinating model interactions.
- Modern React Interface — Lightweight frontend built with React and Vite.
- Express Backend — REST-oriented Node.js backend for application services.
- Environment-Based Configuration — Uses environment variables through
dotenvfor configurable backend behavior. - Experimental Architecture — Built as a playground for testing new AI interaction patterns and application ideas.
- React
- React DOM
- Vite
- JavaScript
- Node.js
- Express
- CORS
- dotenv
- LLM orchestration service
- Model-driven application workflows
boot/
├── client/
│ ├── public/
│ ├── src/
│ │ ├── App.jsx
│ │ ├── App.css
│ │ ├── index.css
│ │ └── main.jsx
│ ├── package.json
│ └── vite.config.js
│
├── server/
│ ├── services/
│ │ └── llmOrchestrator.js
│ ├── server.js
│ └── package.json
│
└── .gitignore
cd client
npm install
npm run devIn another terminal:
cd server
npm install
node server.jsConfigure required environment variables in the server environment before running the application.
BOOT is primarily an experimental AI engineering project.
The objective is to investigate how LLMs can move beyond simple chatbot interfaces and become an integrated reasoning or orchestration layer inside a full-stack application.
The architecture can be extended with additional AI providers, tools, agents, external APIs, persistent storage, authentication, and more advanced orchestration logic.
Experimental Prototype
The project is intended for experimentation, architecture exploration, and rapid prototyping rather than production deployment.
Developed by Sakho115.