README

AI Automation Platform is a 100% free and open-source AI workflow orchestration and media generation studio. Build multi-step AI pipelines on an interactive visual canvas that research, write copy, generate images and videos, and publish through MCP connectors.

Anyone can install and run this on localhost, self-host on a cloud server/VPS, or deploy it as a multi-tenant SaaS model for their own team or clients.

✨ Features

Free for Everyone: No subscription paywalls, no monthly fees, and no artificial execution limits. Every feature is unlocked.
Interactive Visual Canvas: Drag and drop nodes (Text, Image, Video, Deep Research, MCP Connectors) powered by @xyflow/react.
Intelligent Multi-Model Routing: Assign best-fit models per task (Claude for research, OpenAI for images, Gemini for video, DeepSeek for bulk copy) with automatic failover fallback on API rate limits.
Deep Research with Live Sources: Autonomous research nodes crawl web pages, cross-check facts, and return cited reports to downstream nodes.
First-Class Media Studio: Built-in asset library with image color palette analysis, consistent character sheets, and automatic video joining using FFmpeg.
Model Context Protocol (MCP): Native connectors for GitHub, Slack, Google Drive, Notion, HubSpot, and custom MCP servers.
Public Shareable Web Forms: Convert any AI pipeline into a public interactive form (no login required for clients) with custom rate limits.
Bring Your Own Keys (BYOK): Pay AI providers directly at wholesale token rates. No middleman markup.
Bank-Grade Key Security: Each workspace possesses its own encrypted data key using AES-256-GCM. Keys are decrypted only in memory during execution, never returned to browsers, and sanitized from logs and error traces.

🔒 Security & Privacy Notice

Important

No database credentials, secrets, or API keys are stored in this repository.

The application relies solely on your local .env configuration.
All AI provider keys you enter through the web interface are securely encrypted into your private database using AES-256-GCM.
Your personal database dumps and .env files are strictly excluded via .gitignore.

📋 Requirements

PHP: ^8.2 (with extensions: pdo, mbstring, openssl, curl, bcmath, fileinfo)
Composer: ^2.x
Node.js: ^18.x or ^20.x & npm
Database: MySQL 8.0+ or SQLite 3
FFmpeg (Optional, recommended): Used for joining multi-scene video clips.

🚀 Quickstart: Local Installation

1. Clone the Repository

git clone https://github.com/developer-ishrafilkhan/ai-automation.git
cd ai-automation

2. Install Dependencies

composer install
npm install

3. Environment Configuration

Copy the template environment file:

cp .env.example .env

Generate the Laravel application key and Vault master encryption key:

php artisan key:generate
php artisan vault:key

4. Database Setup

Set up your database in .env:

For MySQL:

DB_CONNECTION=mysql
DB_HOST=127.0.0.1
DB_PORT=3306
DB_DATABASE=aiautomation
DB_USERNAME=root
DB_PASSWORD=your_password

For SQLite:

DB_CONNECTION=sqlite

(For SQLite, create the database file if needed: touch database/database.sqlite)

Run migrations:

php artisan migrate

Link public storage for media assets:

php artisan storage:link

5. Start the Application

Run the web development server:

composer run dev

In a separate terminal, start the background queue worker (required to process asynchronous AI runs, video renders, and research steps):

php artisan queue:work --timeout=1000

Open your browser at http://localhost:8000 (or your local web server URL). Register an account, add your API keys under Keys, and start creating workflows!

🌐 Deploying to Production (Server / VPS)

Queue Worker (Supervisor)

Create /etc/supervisor/conf.d/ai-automation-worker.conf:

[program:ai-automation-worker]
process_name=%(program_name)s_%(process_num)02d
command=php /path/to/ai-automation/artisan queue:work --sleep=3 --tries=3 --max-time=3600 --timeout=1000
autostart=true
autorestart=true
stopasgroup=true
killasgroup=true
user=www-data
numprocs=4
redirect_stderr=true
stdout_logfile=/path/to/ai-automation/storage/logs/worker.log

Update and start supervisor:

sudo supervisorctl reread
sudo supervisorctl update
sudo supervisorctl start all

Nginx Configuration

Direct your Nginx root to /path/to/ai-automation/public:

server {
    listen 80;
    server_name yourdomain.com;
    root /path/to/ai-automation/public;

    add_header X-Frame-Options "SAMEORIGIN";
    add_header X-Content-Type-Options "nosniff";

    index index.php;

    charset utf-8;

    location / {
        try_files $uri $uri/ /index.php?$query_string;
    }

    location = /favicon.ico { access_log off; log_not_found off; }
    location = /robots.txt  { access_log off; log_not_found off; }

    error_page 404 /index.php;

    location ~ \.php$ {
        fastcgi_pass unix:/var/run/php/php8.2-fpm.sock;
        fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;
        include fastcgi_params;
    }

    location ~ /\.(?!well-known).* {
        deny all;
    }
}

🏢 Operating as a SaaS Platform

This platform comes out of the box with multi-tenant workspace architecture:

Workspaces: Each registered user gets an isolated workspace with its own encryption key, asset library, flow catalog, and MCP servers.
Team Collaboration: Workspaces support member invitations and role assignments.
Public Form Gateways: Users can generate public endpoints for their AI workflows (/f/{slug}) with rate limiting, allowing them to embed workflows or offer services to clients.
Wholesale Pricing / BYOK: Since users supply their own keys, hosting costs remain purely for compute and storage.

🛠️ Tech Stack

Backend: Laravel 12, PHP 8.2+
Frontend: React 19, Inertia.js 2, TypeScript
Styling: Tailwind CSS 4, Radix UI primitives, Lucide Icons
Workflow Engine: Custom DAG execution engine with @xyflow/react
Database: MySQL / SQLite via Eloquent ORM
Queue: Database Queue with asynchronous step polling

👨‍💻 Author & Maintainer

Maintained and developed by Ishrafil Khan at Larwell Technologies.

📄 License

This project is licensed under the MIT License. You are free to use, modify, distribute, and monetize this project for personal or commercial purposes.

larwell / ai-automation

Package info

Security

Statistics