User Guide

1. Installation & First Launch

1.1 Installation

Installing Alva is straightforward. Double-click the installer file you downloaded and follow the on-screen steps. No administrator rights are required — Alva installs to your user folder automatically.

Once the installation is complete, a desktop shortcut is created so you can launch Alva at any time.

1.2 First Launch — Sign In

When you open Alva for the first time, an onboarding flow walks you through the initial setup. You will be asked to sign in with your ChatGPT credentials. This connects Alva to the AI backend that powers the assistant.

After signing in, Alva automatically verifies that all required dependencies are present on your system.

1.3 Dependencies (Meetings)

To use the Meetings mode (recording and transcription), Alva needs a few additional components:

• Python (portable version) — runs locally, no global install required
• FFmpeg — handles audio processing
• Faster-Whisper — the AI model used for speech-to-text transcription

All of these are installed automatically with a single click from the onboarding screen. A progress indicator shows each component as it downloads and installs.

2. Interface Overview

2.1 Overview

The Alva interface is organized into three main zones:

• Left sidebar — Contains the file explorer and mode navigation. Switch between the four modes (Assistant, Meetings, Documents, Inbox) from here.
• Center area — The main workspace. Displays the chat, active mode content, or document viewer depending on what you are doing.
• Header bar — Shows conversation tabs, action buttons, and indicates which mode you are currently in.

2.2 The 4 Modes

2.3 File Explorer

The left sidebar includes a built-in file explorer that gives the AI context about your work. Key features include:

• Quick access — One-click access to your Desktop, Documents, and Downloads folders.
• Folder navigation — Browse your file system with breadcrumb navigation.
• Real-time search — Type to instantly filter files in the current folder.
• File selection for AI context — Select a folder to give the assistant access to those files. The AI can then read, create, and modify files within it.

3. Assistant Mode (AI Chat)

3.1 Send a Message

Type your request into the input field at the bottom of the screen and press Enter to send. The AI responds in real time, streaming text, code, and actions as they are generated.

The assistant has access to the folder you selected in the file explorer. It can:

• Read existing files and understand their content
• Create new files from scratch
• Modify and update existing files
• Answer questions about your project structure

3.2 Conversation Tabs

You can have multiple conversations open simultaneously. Click the "+" button in the header to start a new conversation tab. Each tab maintains its own independent history.

Drag and drop tabs to reorder them as you like.

3.3 Commands and Settings

Click the "..." menu in the top-right corner to access additional options:

• History — Browse and reopen past conversations
• Settings — Configure AI model, theme, shortcuts, and more
• Support — Access help and support resources

3.4 Skills

Skills are pre-configured instructions that guide the AI for specific tasks. When you select a skill, the assistant receives tailored instructions that improve its output for that particular use case.

Select a skill from the dropdown above the input field before sending your message.

4. Meetings Mode (Transcription)

4.1 Start Recording

Open the floating recording window by pressing Ctrl+Shift+Space or by clicking the recording button in Meetings mode. Then click the red record button to begin.

Alva captures both your system audio (from applications like Teams, Google Meet, Zoom, etc.) and your microphone simultaneously.

4.2 During Recording

A red indicator confirms that recording is active. You can continue using your computer normally — work in other applications, browse the web, or take notes. The recording runs quietly in the background.

4.3 Transcribe

When the meeting is over, click "Transcribe". The AI processes the audio, identifies different speakers, and generates a full text transcript. Once complete, you are redirected to the assistant where you can ask questions about the meeting content.

4.4 Use the Transcription

After transcription, quick action buttons let you instantly generate:

• Summary — A concise overview of the meeting
• Key Points — The most important topics discussed
• Action Items — Tasks and follow-ups assigned during the meeting
• Full Transcript — The complete transcript with timestamps

4.5 Audio Permissions

System audio (from other applications) is captured automatically without additional configuration. However, your microphone requires Windows permission to be accessed.

5. Documents Mode

5.1 Analyze a Document

Drag and drop a file into the Documents view, or select one from the file explorer. Once the document is loaded, type your question in the input field — the AI will analyze the content and respond.

5.2 Supported Formats

Documents mode supports the following file types:

6. Inbox Mode (Mail & Tickets)

6.1 Connect Outlook

To use your Outlook email in Alva, go to Settings > Inbox and enter your email address along with an app password. App passwords are required because Alva uses IMAP to access your mailbox securely.

For step-by-step instructions on generating an app password, see the Outlook section of the Integrations guide.

6.2 Connect Jira

To pull in Jira tickets, go to Settings > Inbox and enter your Jira email, API token, and site URL (e.g., yourcompany.atlassian.net).

For instructions on generating a Jira API token, see the Jira section of the Integrations guide.

6.3 Use the Inbox

Once connected, Inbox mode provides a unified view of your emails and tickets:

• Filter — Switch between All, Mail, and Tickets views
• Browse — Scroll through your messages and tickets in a clean list
• View details — Click any item to see its full content
• Chat with AI — Ask the assistant questions about specific emails or tickets, draft replies, or extract information

7. Settings & Customization

Open Settings from the "..." menu > Customization, or use the keyboard shortcut Ctrl+,.

7.1 Theme

Choose between two interface themes:

• Light (default) — White background, optimized for bright environments
• Dark — Dark background, easier on the eyes in low-light settings

7.2 AI Model

Select which AI model the assistant uses. Different models offer varying trade-offs between speed and capability. Change this setting to experiment with what works best for your tasks.

7.3 Custom Prompt

Define global instructions that are sent with every message to the AI. This is useful for setting:

• A preferred tone or communication style
• The language the AI should respond in
• Specific rules or constraints for responses

7.4 Custom Skills

Create your own skills to automate repetitive tasks. Each custom skill includes:

• Name — A short label for the skill
• Description — What the skill does (shown in the dropdown)
• Detailed instructions — The full prompt sent to the AI when this skill is active
• Prerequisites — Any files or context the skill expects
• Examples — Sample inputs and outputs to guide the AI

7.5 Global Shortcut

The default global shortcut to open the Alva overlay from any application is Ctrl+Shift+Space. You can customize this to any key combination you prefer in Settings.

7.6 Quick Access

Customize which folders appear in the quick access section of the file explorer. By default, Desktop, Documents, and Downloads are shown. You can add any custom folder path.

7.7 Gemini Configuration

Configure the Gemini API for Documents mode:

• API Key — Your Google Gemini API key
• Temperature — Controls response creativity (lower = more focused, higher = more creative)
• Max Tokens — Maximum length of the AI response

See the Integrations guide for instructions on obtaining an API key.

7.8 Meeting Settings

Configure how Meetings mode behaves:

• Language — The language spoken in your meetings (used for transcription accuracy)
• Whisper Model — Choose between small (faster transcription) and large (more accurate transcription)
• Storage Folder — Where recorded audio files and transcripts are saved

8. Keyboard Shortcuts

General

Navigation

Meetings

Global

9. Updates

9.1 Manual Update

To update Alva manually, follow these steps:

1. Close Alva completely — Simply closing the window may not be enough. Open Task Manager (press Ctrl+Shift+Esc), find the process named "Electron" or "Alva", and click End Task.
2. Download the latest installer — Get the new version from the download page.
3. Run the installer — It installs over the existing version. Your data, settings, conversation history, and custom skills are all preserved.
4. Launch Alva — Open the app normally. You are now on the latest version.

10. FAQ & Troubleshooting