Stop Guessing. Start Forecasting.
Fyller is a Next-Gen Predictive AI Sales Forecasting & Inventory System that learns your sales patterns, predicts demand, and automates restocking. Transform your inventory from a liability into a strategic asset. Unlike standard CRUD inventory apps, this system implements Double-Entry Ledger Accounting principles for 100% stock accuracy and leverages Generative AI to predict future demand based on historical sales trends.
- Predictive Sales Forecasting: AI analyzes historical data and seasonal trends to predict future sales with 98% accuracy.
- Real-Time Sync: Instant inventory updates across web, mobile, and physical POS.
- Smart Safety Stock: Auto-adjustment of stock levels based on supplier lead times and volatility.
- Actionable Insights: Recommendations for bundling, discounting, and restocking.
- CFO-Level Dashboard: Visualize revenue trends, identify slow-moving stock, and generate AI purchase orders.
- Framework: Laravel 12 (PHP 8.2+)
- Frontend: React 19, Inertia.js 2.0
- Styling: Tailwind CSS 4, Framer Motion
- Icons: Lucide React
- Build Tool: Vite
- PHP 8.2+
- Composer
- Node.js & NPM
We follow a strict Domain-Driven structure to ensure scalability:
app/
├── Domain/
│ ├── Inventory/ # Stock logic, Ledgers, Products
│ ├── Sales/ # Orders, Customers, Invoicing
│ └── Forecasting/ # AI logic, Daily Sales Aggregation
├── Http/
│ └── Controllers/ # API & Web Entry Points
└── Infrastructure/ # External Integrations (OpenAI, AWS)
-
Clone the repository
git clone <repository-url> cd fyller
-
Install dependencies
composer install npm install
-
Create a .env file
cp .env.example .env
-
Generate an application key
php artisan key:generate
-
Update your .env with your database credentials and OpenAI API Key:
DB_CONNECTION=pgsql OPENAI_API_KEY=sk-your-key-here
-
Run migrations
php artisan migrate
-
Run the development server
composer run dev
Visit http://127.0.0.1:8000/ to access the dashboard.
Run the test suite with:
```bash
composer run test
```
Fyller is open-sourced software licensed under the MIT license.