Brahma AI - Lite

Open-source Windows desktop AI assistant for voice, automation, productivity, and intelligent workflows.

---

Overview

Brahma AI - Lite is an open-source desktop assistant designed for Windows power users. It unites voice and text input with intelligent automation, productivity workflows, document generation, and adaptive screen-aware actions.

• Live voice and text interaction via Gemini with OpenRouter fallback
• Automatic daily briefing with interruption-aware audio playback
• Desktop automation for apps, windows, files, and browser workflows
• Office content generation for PowerPoint, Word, spreadsheets, and PDF
• Built-in website and workspace creation through Brahma's local workspace generator
• Discord collaboration, reminders, meeting assistant, and notifications

Why It Stands Out

• Responsive UI with rich task/workspace feedback
• Seamless AI and voice integration for desktop productivity
• Modular tool-driven architecture for clean extension
• Robust fallback handling to keep the assistant available
• Local-first configuration with secure credentials storage

Features

Intelligent Assistant

• Unified voice + typed conversation experience
• Startup daily briefing with Edge TTS delivery
• Instant briefing interruption when a new message arrives
• Gemini-first live AI with OpenRouter fallback support

Automation & Productivity

• Open and control Windows applications and system actions
• Browser automation with Playwright for web workflows
• Screen inspection and contextual content extraction
• File and document automation for fast productivity

Office & Content Tools

• Generate PowerPoint decks and presentation content
• Create spreadsheets and Word documents quickly
• Export polished reports as PDF files
• Build landing pages and websites from within the app

Integrations

• Discord bridge for remote commands and chat
• Local credential management for Gemini and OpenRouter
• Configurable voice, notifications, startup, and UI preferences

How It Works

Brahma AI - Lite is built on a layered desktop architecture that separates UI, AI session management, and tool execution.

• main.py initializes the application, launches the UI, and manages the AI runtime.
• BrahmaLive owns the live AI session, audio queues, and command routing.
• actions/ contains modular tools for automation, document generation, notifications, meetings, and search.
• AttentionMonitor captures external events and notification text, then speaks alerts using Edge TTS.
• daily_briefing.py constructs the morning briefing text and triggers playback after startup.
• ui.py provides a polished Qt-based interface with command entry, workspace cards, and status feedback.
• or_client.py provides OpenRouter fallback support when Gemini is unavailable or rate-limited.

Getting Started

Prerequisites

• Windows 10 or Windows 11
• Python 3.11 or 3.12
• Git installed
• Gemini API key
• OpenRouter API key (optional, recommended for fallback resilience)

1. Clone the repository

git clone https://github.com/titechprabhasolutions/Brahma-AI---Lite.git
cd "Brahma AI - Lite"

2. Create a virtual environment

python -m venv .venv
.\.venv\Scripts\Activate.ps1

3. Install dependencies

pip install -r requirements.txt
playwright install

4. Configure API keys

The app loads keys from config/api_keys.json. Create this file if it does not already exist.

{
  "gemini_api_key": "YOUR_GEMINI_API_KEY",
  "openrouter_api_key": "YOUR_OPENROUTER_API_KEY"
}

Gemini API Key

• Create a Google Cloud or Gemini account
• Enable Gemini API access for your project
• Generate an API key and add it under gemini_api_key

OpenRouter API Key (recommended)

• Register at https://openrouter.ai
• Generate an sk-or- API key
• Add it under openrouter_api_key

5. Optional: Configure Discord integration

To enable Discord bridging, populate config/discord_bot.json with your bot credentials and connection settings.

6. Start the app

python main.py

For a cleaner Windows launch without console output:

start_brahma.vbs

Configuration

• config/api_keys.json — Gemini and OpenRouter credentials
• config/app_settings.json — voice, UI, startup, and automation settings
• config/discord_bot.json — Discord bridge settings

Plugin System

You can extend Brahma AI with lightweight Python plugins placed in the plugins/ folder. Plugins are simple *.py files that export one or more hook functions:

• on_brahma_created(brahma) — called when the BrahmaLive instance is created
• on_startup(brahma) — called once during app startup after plugins are registered
• on_text_command(text, source, brahma=None) — called for each incoming text command; return True to mark the command handled and stop further processing

Example: plugins/example_plugin.py demonstrates the hooks.

To enable plugins, simply drop your plugin file into the plugins/ folder and restart the app. The launcher and main.py load plugins automatically.

Project Structure

• main.py — core runtime, AI session orchestration, and startup flow
• ui.py — polished Qt interface, workspace cards, and controls
• actions/ — modular tools for automation and AI workflows
• config/ — local settings, API keys, and runtime configuration
• tests/ — validation and integration tests for core features

Recent Updates

2026-07-19

• Restored automatic daily briefing playback at startup.
• Unified local TTS output to the same Edge voice.
• Added briefing interruption support for immediate user response.
• Redesigned gesture HUD for hand landmark control.
• Improved cursor mapping for better desktop reach and direction.
• Added idle speech prompts for proactive engagement.
• Tightened the developer workflow so Brahma uses the local workspace generator for websites and keeps coding tasks in the main app flow.

Community

• Discord: https://discord.gg/gEYmJKKtq3

License

This project is licensed under a custom source-available license. See LICENSE for full terms and TRADEMARK.md for branding details.

Maintained by

• Suryaansh Tiwari

Please preserve attribution and keep credentials secure when building on top of Brahma AI - Lite.