Intro Snippet: Hey Dev Community! I've spent the last few months building a personal assistant that doesn't just chat—it acts.
Today, I’m open-sourcing Ved AI. It’s a local-first autonomous agent that uses Gemini Pro for reasoning and LiveKit for lightning-fast voice interactions.
Why I built this: Most assistants today feel like silos. Ved is designed to reach into your workflow—Gmail, Drive, System Control—and get things done without your data leaving your control.
The Tech Stack:
Backend: FastAPI & Python
AI: Google Gemini & Groq
Voice: LiveKit Realtime API
Memory: Knowledge Graph (NetworkX)
I’d love for you guys to check out the source code, play with the "Setup-First" UI, and let me know your thoughts!
Ishaan-Sharma-tech
/
ved-ai
Next-generation Autonomous AI Assistant with Real-time Multi-modal Voice, Agentic Swarms, and full Google Workspace integration. Built for Performance & Privacy.
🚀 Ved AI: Your Advanced Autonomous Personal Agent
Ved is a next-generation AI assistant designed to handle your digital life through voice and text. Built with a "Privacy-First" approach, Ved runs locally and connects to cutting-edge AI models to provide a seamless, sharp, and highly capable agentic experience.
🌟 Key Features
- 🎙️ Real-time Multi-modal Voice: Powered by LiveKit and Gemini Realtime API for natural conversations.
- 🐝 Agentic Swarms: Spawns parallel agents for Research, Synthesis, and Data Retrieval.
- 🛠️ Integrated Toolset: Full control over your local System, Gmail, Google Drive, Google Tasks, YouTube, and the Web.
- 📱 Telegram Integration: Secure remote control of your agent from anywhere.
- ⚙️ Setup-First UI: Configure everything (Identity, Keys, Paths) directly from the Web UI.
⚡ One-Click Quick Start (Windows)
The easiest way to get started is using our automated launcher:
-
Clone the repo:
git clone https://github.com/your-username/ved-ai.git -
Double-click
start_ved.bat: This will automatically- Create your
.env…
- Create your
Top comments (0)