🤖 AI Speech-Based Agent Project: Automate Setup of uv, PyAudio, and SpeechRecognition

Automate Python Environment Setup for Real-Time Speech Recognition and TTS

🤖 AI Speech-Based Agent Project: (I) Automate Setup of whisper.cpp

🤖 AI Speech-Based Agent Project: (II) Automate Setup of llama.cpp

🤖 AI Speech-Based Agent Project: Automate Setup of uv, PyAudio, and SpeechRecognition

🧩 Introduction

If you’re building an AI voice assistant, smart speaker, or chatbot that listens and talks, your Python project needs a reliable setup that includes:

• Microphone and speaker access (PyAudio)
• Speech-to-text capabilities (SpeechRecognition)
• Text-to-speech output (edge-tts)
• A fast and modern dependency manager (uv)

Manually setting all this up every time is repetitive and error-prone. In this tutorial, we’ll automate the entire setup using a simple Bash script.

❓ Why These Packages?

Here’s why the key packages are used:

These tools give your AI agent the ability to hear, understand, and speak.

💡 Why an Automatic Bash Script?

Instead of manually running 6–10 shell commands every time:

Reproducibility and ease of onboarding are key for any AI project.

Instead of manually:

• Installing system packages
• Creating a virtual environment
• Installing Python dependencies
• Automate installation for consistency
• Avoid forgetting dependencies
• Ensure uv and environment setup are done properly
• Make it easy to onboard new team members or rebuild your environment later

You can automate it all with a single script that runs in seconds. This ensures:

• 🔁 Consistent environments
• 🧪 Fewer setup mistakes
• 👥 Faster onboarding for teammates

🛠️ Step-by-Step: Build the Bash Script

Let’s break the script into parts so you understand how it works.

✅ Step 1: Script Header

#!/bin/bash
set -e

This tells the system to run the script using Bash and stop if any command fails.

✅ Step 2: Configuration

PROJECT_DIR="voice_project"
PYTHON_VERSION="python3.10"

Customize your project folder and desired Python version.

✅ Step 3: Install System Dependencies

sudo apt update
sudo apt install -y portaudio19-dev

Install the required audio library (portaudio) to enable microphone access through PyAudio.

✅ Step 4: Ensure uv is Installed

if ! command -v uv &> /dev/null; then
    curl -Ls https://astral.sh/uv/install.sh | bash
    export PATH="$HOME/.cargo/bin:$PATH"
fi

If uv isn’t installed, download and install it. This will let you manage dependencies efficiently.

✅ Step 5: Create Project Directory

if [ ! -d "$PROJECT_DIR" ]; then
    mkdir "$PROJECT_DIR"
fi
cd "$PROJECT_DIR"

Create your project folder (if needed) and switch into it.

✅ Step 6: Set Up the Virtual Environment

uv .venv --python "$PYTHON_VERSION"
source .venv/bin/activate

Create and activate a virtual environment using uv.

✅ Step 7: Add Required Packages

uv add pyaudio SpeechRecognition edge-tts

Install the core packages your AI agent needs to listen and speak.

✅ Step 8: Export a Reproducible Requirements File

uv export --without-hashes > requirements.txt

Create a requirements.txt so others (or CI/CD) can replicate the environment with pip.

📄 Full Script: install.sh

#!/bin/bash
set -e

PROJECT_DIR="voice_project"
PYTHON_VERSION="python3.10"
echo "📦 Installing system dependencies..."
sudo apt update
sudo apt install -y portaudio19-dev
echo "🧪 Checking if uv is installed..."
if ! command -v uv &> /dev/null; then
    echo "🔧 Installing uv..."
    curl -Ls https://astral.sh/uv/install.sh | bash
    export PATH="$HOME/.cargo/bin:$PATH"
fi
echo "📁 Creating project directory if not exists..."
if [ ! -d "$PROJECT_DIR" ]; then
    mkdir "$PROJECT_DIR"
fi
cd "$PROJECT_DIR"
echo "🌀 Creating virtual environment..."
uv .venv --python "$PYTHON_VERSION"
echo "🚀 Activating virtual environment..."
source .venv/bin/activate
echo "📦 Installing Python packages..."
uv add pyaudio SpeechRecognition edge-tts
echo "📄 Exporting requirements.txt..."
uv export --without-hashes > requirements.txt
echo "✅ Setup complete! You can now start coding."

🚀 How to Use

Follow these simple steps to automatically set up your AI speech-based project:

1. ✅ Save the Script

Create a new file named voice_recong_tts_install.sh and paste in the full script:

nano voice_recong_tts_install.sh
# Or use your preferred editor

Make the script executable:

chmod +x voice_recong_tts_install.sh

2. 🏃‍♂️ Run the Script

Run the script from your terminal:

./voice_recong_tts_install.sh

This will:

• Install system dependencies (portaudio19-dev)
• Install uv if missing
• Create a project folder (e.g., voice_project)
• Set up a virtual environment using uv
• Install pyaudio, SpeechRecognition, edge-tts
• Export a requirements.txt file

3. 🔁 Activate the Environment (Later Sessions)

Each time you return to your project, activate the environment manually:

cd voice_project
source .venv/bin/activate

4. 🧪 Test the Setup

🎤 STT Test: stt_test.py

You can now create a stt_test.py to test your microphone and speech-to-text:

import speech_recognition as sr

# Initialize recognizer
recognizer = sr.Recognizer()

# Use the default microphone as the audio source
with sr.Microphone() as source:
    print("🎙️ Say something...")
    recognizer.adjust_for_ambient_noise(source)  # Optional: better accuracy
    audio = recognizer.listen(source)

print("🔊 Recognizing...")

# Try to recognize speech using Google Web Speech API
try:
    text = recognizer.recognize_google(audio)
    print("✅ You said:", text)
except sr.UnknownValueError:
    print("❌ Could not understand audio.")
except sr.RequestError as e:
    print(f"⚠️ Could not request results; {e}")

▶️ How to Run

Make sure your virtual environment is activated:

source .venv/bin/activate

Then run:

python stt_test.py

Speak clearly into your microphone when prompted.

🔊 TTS Test: tts_test.py

import asyncio
from edge_tts import Communicate

async def main():
    text = "Hello! I am your AI speech agent. This is a text-to-speech test."
    voice = "en-US-AriaNeural"  # You can change this to any available Microsoft voice
    communicate = Communicate(text, voice)
    await communicate.save("tts_output.mp3")
    print("✅ Speech saved to tts_output.mp3")
asyncio.run(main())

▶️ How to Run

python tts_test.py

Then play the generated file:

# Use your system's audio player
ffplay tts_output.mp3       # If you have ffmpeg
# or
mpg123 tts_output.mp3       # On some Linux systems
# or just double-click it in your file explorer

✅ Summary

With this one-click voice_recong_tts_install.shscript, you now have:

• A full project directory with a virtual environment
• Leverage uv for a fast and modern Python development workflow
• All dependencies installed (pyaudio, speechrecognition, edge-tts)
• Set up a clean Python project with audio and speech capabilities
• A reproducible requirements.txt
• A clean start for building your AI speech-based assistant
• Avoid manual setup errors

This script is perfect for projects like voice assistants, speech-to-text apps, or conversational bots.