Basic Usage

Sign In

When you open ExpellusAI, you will see the sign-in screen. Use the same account you created on the ExpellusAI website (expellusai.com) to log in. You can sign in with your email and password, or use "Continue with Google" for quick access.

Main Window Overview

After signing in, you will see the main window. It has four areas:

• Top-left — Your Account: Shows your profile icon and account information.
• Top-right — AI Query Box: This is where you type what you want the AI to do for you.
• Bottom-left — Action Palette: A collection of action templates (like Click, Type, Scroll, etc.) that you can drag onto the canvas to build or customize workflows.
• Bottom-right — Canvas: Your workspace. This is where the AI builds a visual step-by-step plan of the actions it will take on your screen.

You will also notice a small < button on the right edge of the screen. Clicking it opens the Side Panel — a slide-out panel where you can also type instructions and see a live log of what the AI is doing.

User Settings

Click the user icon in the top-left corner to open the User Settings menu. From here, you can access:

• Change Profile: Opens your account settings page on the ExpellusAI website, where you can update your profile.
• Personal Information: Opens a window on the canvas where you can securely store private details the AI may need (explained in the next section).
• Service Context: Opens a window on the canvas showing the software installed on your computer (explained in Section 5).
• Key File Location: Opens the folder on your computer where ExpellusAI stores your local settings and data files.

Personal Information (Private Credential Store)

The Personal Information window lets you store sensitive details — like usernames, passwords, addresses, or any form-fill data — that the AI might need to complete tasks on your behalf.

Here is how it works:

• Each entry has three fields: Attribute Name (e.g., "Work Email"), Remarks(optional extra context), and Value(the actual sensitive data, like your email address).
• The AI can see the Attribute Name andRemarks so it knows what kind of information is available, but it never sees the Valueitself. The Value is only applied locally on your computer at the exact moment it is needed — for example, when filling in a login form.
• All data is saved locally on your device (in the Key File Location folder) and never uploaded to any server. Only the Attribute Names and Remarks are shared with the AI to provide context.

Service Context (Installed Software)

The Service Context window tells the AI what software you have installed on your computer. This helps the AI understand which tools are available when performing tasks (for example, whether to open Chrome, Firefox, Notepad, or Excel).

• By default, the list is empty. Click the auto-detect button(the icon next to the + button) to automatically scan and import your installed applications.
• You can also manually add or removeentries.

Starting a Task

To give the AI a task, simply type your instruction in the AI Query Box at the top of the main window. For example, you could type:

• "Open Chrome and search for today's weather"
• "Write my IP address in Notepad"
• "Fill out the registration form on this website"

Press Enter or click the arrow button (→) to start. The AI will immediately begin working on your request.

You can also start a task from the Side Panel — just open it using the < button on the right edge, and use the input field at the bottom.

While the AI Is Working

Once the AI starts working, it will take control of your mouse and keyboard to perform actions on screen. During this time:

• The Side Panel will automatically appear and disappear as the AI takes actions, showing you a real-time log of what it is doing.
• Your main window will minimize to the taskbar while the AI works, and will come back automatically when the task is finished.

Keyboard shortcuts (you can customize these in Settings > Preferences):

When the task is complete — or if you force‑stop it — the Side Panel will display an "End of Iteration" message, and the AI will stop.

Reviewing Results on the Canvas

After the AI finishes, the main window returns and shows a visual record of every action it performed. Each action is displayed as a block on the canvas, connected in the order they were executed.

You can:

• Re‑run the entire flow by clicking the Execute button in the top‑right corner. This replays all the actions exactly as they were performed.
• Skip specific AI steps — Each "AI Request" block has a Skip toggle. When enabled, the AI will not be called for that step during re‑execution, saving token costs.
• Connect blocks — Drag from the small white circle on one block to another to define the execution order. This lets you rearrange or customize the flow.

Editing and Saving Flows

You can fine‑tune any action the AI performed:

• Double‑click any action block to edit its parameters — for example, changing click coordinates, modifying the text to type, or adjusting a scroll amount.
• Resize blocks by dragging the dashed handle at the bottom‑right corner of any block.

You can also save your workflows as project files and reload them later:

• Save — Saves the current canvas layout and all action blocks to a file.
• Load — Opens a previously saved project.
• New — Clears the canvas to start fresh.

This means you can build reusable automation flows — create them once, save them, and re‑run them anytime.