Conversational Audio Bot

An AI-powered voice and text customer service assistant built with Streamlit, OpenAI, and LangChain.

Features

• 🎤 Voice Input - Record audio or upload audio files (WAV/MP3)
• ⌨️ Text Input - Type messages directly
• 🔊 Text-to-Speech - AI responses are converted to audio
• 💬 Conversation History - Full chat history with context awareness
• 📊 Conversation Analysis - AI-powered analysis of customer interactions
• 🎯 Multiple Modes - Customer Service and Lead Generation personas

Quick Start

Prerequisites

• Python 3.10+
• Poetry (recommended) or pip
• OpenAI API key
• PortAudio (for microphone recording on Linux)

Installation

1. Clone the repository

   git clone <repository-url>
   cd conversational-audio-bot

2. Install system dependencies (Linux only - for microphone support)

   sudo apt-get update && sudo apt-get install -y portaudio19-dev python3-dev

3. Install Python dependencies

   Using Poetry:

   poetry install
   poetry shell

   Or using pip:

   python -m venv .venv
   source .venv/bin/activate  # Linux/Mac
   # .venv\Scripts\activate   # Windows
   pip install -e .

4. Configure environment variables

   Create a .env file in the project root:

   OPENAI_API_KEY=your-openai-api-key-here
   SAMPLE_RATE=16000
   LLM_MODEL=gpt-4

5. Run the application

   streamlit run app.py

   The app will open at http://localhost:8501

Usage

Voice Interaction

1. Go to the Voice Interaction tab
2. Click Start Recording to record from your microphone (requires audio device)
3. Or use Upload audio file to upload a WAV/MP3 file
4. The audio is transcribed and the AI responds with text and audio

Text Interaction

1. Go to the Text Interaction tab
2. Type your message and press Enter
3. The AI responds with text and audio

Conversation Analysis

1. Have a conversation first
2. Go to the Conversation Analysis tab
3. Click Analyze Conversation to get AI insights

Settings (Sidebar)

• Recording Duration - Adjust recording length (5-20 seconds)
• TTS Language - Choose English or Chinese
• Mode - Switch between Customer Service and Lead Generation
• Clear History - Reset conversation

Troubleshooting

"Recording failed" / "Error querying device -1"

This means no microphone is available. Common in WSL or remote environments.

• Solution: Use the file upload feature or text input instead

"Import could not be resolved" errors

Run poetry install or pip install -e . to install dependencies.

OpenAI API key errors

Ensure your .env file has OPENAI_API_KEY (not OPEN_AI_KEY).

Project Structure

conversational-audio-bot/
├── app.py                 # Main Streamlit application
├── utils/
│   ├── audio_utils.py     # Audio recording, processing, TTS
│   └── llm_utils.py       # LLM integration with LangChain
├── pyproject.toml         # Project dependencies (Poetry)
├── .env                   # Environment variables (create this)
├── README.md              # This file
└── ARCHITECTURE.md        # Detailed architecture documentation

Architecture

See ARCHITECTURE.md for detailed technical documentation including:

• System architecture diagram
• Component details
• Data flow
• Configuration options

License

See LICENSE for details.