Skip to content

1.4. First-Time Setup Wizard

When you launch Backend.AI GO for the first time, you'll be guided through an interactive setup wizard (also called the OOBE - Out-of-Box Experience). This wizard helps you configure essential settings before using the application.

Overview

The setup wizard consists of four steps:

  1. Language Selection - Choose your preferred interface language
  2. Engine Selection - Select the inference engine for running models
  3. Models Directory - Configure where models are stored
  4. Setup Complete - Review your settings and start using the app

Step 1: Language Selection

Choose the language for the Backend.AI GO interface.

Currently supported languages:

  • English - Full support
  • 한국어 (Korean) - Full support

Click on your preferred language card to select it. The interface will immediately switch to the selected language.

Changing Language Later

You can change the language at any time in Settings → General → Language.

Step 2: Engine Selection

Select the inference engine that will run your AI models. The wizard automatically detects your hardware capabilities and recommends the best option.

Let Backend.AI GO automatically select the best engine for each model based on:

  • Your hardware capabilities
  • The model format (GGUF, MLX, etc.)
  • Optimal performance settings

Manual Mode

Choose a specific engine to always use:

Engine Platform Best For
llama.cpp All platforms GGUF models, GPU acceleration via Metal/CUDA/ROCm
MLX macOS (Apple Silicon) Apple Silicon optimization, native performance
vLLM Linux High-throughput serving, advanced batching

Hardware Detection

The wizard displays your system's capabilities:

  • GPU Type: Metal (macOS), CUDA (NVIDIA), ROCm (AMD), or CPU only
  • VRAM: Available video memory for model loading
  • Supported Engines: Engines compatible with your system

Step 3: Models Directory

Configure the location where downloaded models will be stored.

Default Location

By default, models are stored in:

Platform Default Path
macOS ~/Library/Application Support/backend.ai-go/models/
Windows %APPDATA%\backend.ai-go\models\
Linux ~/.local/share/backend.ai-go/models/

Custom Location

Click Browse to select a different folder. Consider:

  • Available Space: Large models can be 4-20+ GB each
  • Drive Speed: SSDs provide faster loading times
  • External Storage: You can use external drives if needed

Storage Requirements

Make sure your selected location has sufficient free space. A typical setup with 3-5 models may require 30-50 GB.

Headless Mode (Web Browser)

When accessing Backend.AI GO through a web browser in headless server mode, native file dialogs are not available. Instead, clicking Browse opens a server-side directory browser that lets you navigate the server filesystem and select a directory.

The directory input field is also pre-populated with the server's default models directory so you can accept the default or type a custom path directly.

Step 4: Setup Complete

Review your configuration before starting:

  • Language: Your selected interface language
  • Engine Mode: Automatic or manual selection
  • Selected Engine: If manual mode, the chosen engine
  • Models Directory: Where models will be stored

Click Get Started to complete the setup and enter the main application.

  • Use Back to return to previous steps and modify settings
  • Use Next to proceed to the following step
  • Keyboard shortcuts:
    • Left - Go back
    • Right - Go next
    • Enter - Complete setup (on final step)

After Setup

Once the wizard completes, you'll be taken to the Getting Started page where you can:

  • Quick Start: Access primary features (Chat, Models, Agent)
  • Usage Stats: View your usage statistics
  • Featured Models: Discover and download popular models

Onboarding Tour

After completing the first-time setup wizard, an interactive guided tour will automatically start to help you learn about the key features of Backend.AI GO.

Tour Steps

The tour highlights the following areas:

  1. Welcome - Introduction to the tour
  2. Models - Browse and manage your AI models
  3. Chat - Have conversations with AI models
  4. Engines - Manage inference engines
  5. Settings - Configure your preferences
  6. Keyboard Shortcuts - Learn about keyboard navigation
  7. Complete - You're ready to go!

Tour Controls

  • Next - Proceed to the next step
  • Back - Return to the previous step
  • Skip Tour - Exit the tour at any time
  • Get Started - Complete the tour (on final step)

Replaying the Tour

If you want to take the tour again:

  1. Navigate to Settings → UI
  2. Find the Help & Onboarding section
  3. Click Start Tour to replay the tour

What's New Modal

When you update to a new version of Backend.AI GO, a "What's New" modal will appear showing:

  • Highlights - Key improvements in the new version
  • New Features - Newly added capabilities
  • Improvements - Enhancements and fixes

From this modal, you can choose to take the tour to explore new features, or close it to continue using the app.

Resetting the Setup

If you need to run the setup wizard again:

  1. Navigate to Settings → General
  2. Find the Reset First-Time Setup option
  3. Click to reset, then restart the application

This is useful if you want to reconfigure your engine or models directory from scratch.