🤖 BYTE

Your autonomous, personal AI agent. Tell it what to do — it gets it done.

Byte AI dynamically handles complex tasks including real-world voice calls, deep web research, and personal scheduling. Built using a modern React 19 + Node.js stack with Gemini & ElevenLabs integration.

🌟 Key Features

• 🗣️ Voice Execution: Dispatches AI agents to handle real-world phone calls (bookings, inquiries, negotiations) via Twilio and ElevenLabs.
• 🧠 Intelligent Orchestration: Dynamically combines web search and task execution based on user intent.
• 🛡️ Social Interaction Buffer: Intentionally handles anxiety-inducing, human-facing tasks (like calling customer service) to act as a buffer for introverted users or those facing social exhaustion.
• 🎯 Contextual Personalization: Maintains a persistent memory that adapts to user preferences and past interactions for highly tailored execution.

---

🏗️ High-Level Architecture

---

📊 Feature Workflows

1. The Autonomous Orchestration Flow

Byte breaks down complex prompts into step-by-step actions, orchestrating the right tools for the job.

graph TD
    User([User Prompt]) --> Orchestrator
    Orchestrator{Orchestrator Agent}
    Orchestrator -->|Needs Info?| WebSearch[Web Search Tool]
    Orchestrator -->|Needs to Call?| VoiceDispatch[Voice Agent Dispatch]
    Orchestrator -->|Needs to Remember?| Memory[Update User Context]
    WebSearch --> Orchestrator
    VoiceDispatch --> Orchestrator
    Orchestrator --> UI(Real-time Status to UI)

Loading

2. Social Friction Buffer: The Call Execution Pattern

How Byte handles nerve-wracking phone calls on your behalf.

---

🛠️ Tech Stack

Frontend: React 19, Vite, Tailwind CSS 4, HeroUI, Zustand
Backend: Node.js, Express, TypeScript, Zod
AI/Voice Engine: Google Gemini, Groq, ElevenLabs, Twilio
Real-time & DB: WebSockets (Socket.io), MongoDB + Mongoose

---

🚀 Setup & Installation

Prerequisites

• Node.js (v22+)
• MongoDB (running locally or a connection string)
• API Keys: Google Gemini, ElevenLabs, Twilio, Serper (Search API)

1. Clone & Install Dependencies

From the root directory:

# Install frontend dependencies
cd frontend
npm install

# Install backend dependencies
cd ../server
npm install

2. Environment Configuration

Copy the example environment file in the server directory and fill in your API keys:

cd server
cp .env.example .env

Your .env should look something like this:

PORT=3001
JWT_SECRET=your_super_secret_jwt_key

GEMINI_API_KEY=your_gemini_api_key

# Database Connections
SPACETIMEDB_HOST=localhost:3000
SPACETIMEDB_MODULE=byte-dev

# Telephony integration
TWILIO_ACCOUNT_SID=ACxxxxxxxxxxxxxxxx
TWILIO_AUTH_TOKEN=your_twilio_auth_token
TWILIO_PHONE_NUMBER=+1xxxxxxxxxx

# Voice Model
ELEVENLABS_API_KEY=your_elevenlabs_api_key
ELEVENLABS_AGENT_ID=your_elevenlabs_agent_id

# Web search capabilities
SEARCH_API_KEY=your_serper_dev_api_key

# Required for webhooks (Twilio callback)
SERVER_BASE_URL=https://your-ngrok-url.ngrok.io

3. Run the Application

Start both the backend server and the frontend client simultaneously.

# Terminal 1: Backend
cd server
npm run dev

# Terminal 2: Frontend
cd frontend
npm run dev

Open your browser and navigate to http://localhost:5173.

---

💡 Usage

1. Sign In / Create Account: Establish your secure local profile.
2. Setup Preferences: Personalize the bot with your preferred tones, default payment info context, etc.
3. Execute Prompts: Use the dynamic interface to issue complex multi-step commands. Just tell it what you need to accomplish, and watch the real-time execution log!