Voice Assistant in Python

Skills & Concepts You'll Gain

How to use Speech Recognition to convert voice to text
How to use Text-to-Speech (TTS) to make Python speak
How to use Python's datetime and webbrowser modules
How to build a command-based automation system
Foundation for building AI assistants, smart home control, and automation tools

Tech Used: Python 3, SpeechRecognition library, pyttsx3 library, datetime module, webbrowser module

Step 1

Install Required Libraries

Open your terminal (Command Prompt on Windows) and run these commands one by one:

Terminal / CMD

pip install SpeechRecognition pyttsx3 pyaudio

If pyaudio fails to install on Windows, run these two commands instead:

pip install pipwin
pipwin install pyaudio

After installing, create a new file called voice_assistant.py in a folder of your choice.

Step 2

Understand the Libraries

Before writing code, understand what each library does:

speech_recognition - Captures audio from your microphone and sends it to Google's speech API to convert it into text
pyttsx3 - Text-to-Speech engine. Makes Python speak out loud using your computer's built-in voice engine. Works offline!
datetime - Built-in Python module to get the current date and time
webbrowser - Built-in Python module to open URLs in your default browser

Step 3

Write the Code

Copy this complete code into your voice_assistant.py file:

voice_assistant.py

# Voice Assistant - NKDevSpace Tutorial

import speech_recognition as sr
import pyttsx3
import datetime
import webbrowser

# --- Initialize the text-to-speech engine ---
engine = pyttsx3.init()
engine.setProperty('rate', 160)   # Speed of speech (words per minute)
engine.setProperty('volume', 1.0) # Volume: 0.0 to 1.0

def speak(text):
    """Make the assistant speak out loud."""
    print(f"Assistant: {text}")
    engine.say(text)
    engine.runAndWait()

def take_command():
    """Listen to the microphone and return the spoken text."""
    recognizer = sr.Recognizer()

    with sr.Microphone() as source:
        print("\nListening... (speak now)")
        recognizer.pause_threshold = 1  # Wait 1 second of silence before stopping
        audio = recognizer.listen(source)

    try:
        print("Recognizing...")
        command = recognizer.recognize_google(audio)
        print(f"You said: {command}")
        return command.lower()
    except sr.UnknownValueError:
        speak("Sorry, I could not understand. Please say that again.")
        return ""
    except sr.RequestError:
        speak("Sorry, I am having trouble connecting. Check your internet.")
        return ""

def run_assistant():
    """Process the voice command and perform the right action."""
    command = take_command()

    if not command:
        return

    if "time" in command:
        current_time = datetime.datetime.now().strftime("%I:%M %p")
        speak(f"The current time is {current_time}")

    elif "date" in command:
        today = datetime.datetime.now().strftime("%B %d, %Y")
        speak(f"Today's date is {today}")

    elif "open google" in command:
        speak("Opening Google")
        webbrowser.open("https://www.google.com")

    elif "open youtube" in command:
        speak("Opening YouTube")
        webbrowser.open("https://www.youtube.com")

    elif "open github" in command:
        speak("Opening GitHub")
        webbrowser.open("https://www.github.com")

    elif "hello" in command or "hi" in command:
        speak("Hello! I am your voice assistant. How can I help you?")

    elif "your name" in command:
        speak("I am NKBot, your personal voice assistant built with Python.")

    elif "stop" in command or "exit" in command or "bye" in command:
        speak("Goodbye! Have a great day.")
        exit()

    else:
        speak(f"I heard you say: {command}. I don't know how to do that yet.")

# ── Main loop ──
speak("Voice Assistant is ready. Say a command.")

while True:
    run_assistant()

Step 4

Code Explanation

engine = pyttsx3.init()

Creates the text-to-speech engine. setProperty('rate', 160) controls how fast it speaks - lower number = slower speech. You can adjust this to your preference.

engine.say(text) → engine.runAndWait()

say() queues the text to be spoken. runAndWait() actually plays the speech and waits until it finishes before continuing the program. Always use both together.

recognizer.recognize_google(audio)

Sends the recorded audio to Google's free Speech-to-Text API over the internet and returns the recognized text. This is why you need an internet connection. The result is converted to lowercase with .lower() so commands like "Open Google" and "open google" both work.

except sr.UnknownValueError

This catches the error when Google couldn't understand what was said (too quiet, background noise, etc.). Instead of crashing, we politely ask the user to repeat.

if "time" in command

We use in to check if a keyword exists anywhere in the command. So "what is the time" and "tell me the time" both match because both contain the word "time".

while True: run_assistant()

The assistant keeps listening in an infinite loop. The only way to stop it is to say "stop", "exit", or "bye" - which calls exit() to end the program.

Step 5

Run the Program

Open your terminal and navigate to your project folder:
CMD / Terminal
```
cd path\to\your\folder
```
For example, if your file is on the Desktop:
CMD
```
cd C:\Users\YourName\Desktop
```
Run the program:
CMD / Terminal
```
python voice_assistant.py
```
You will hear "Voice Assistant is ready. Say a command."
Try saying: "What is the time" - it should speak the current time
Try: "Open YouTube" - your browser should open YouTube
Say "bye" to stop the assistant

Speak clearly and at a normal pace. If it keeps saying "I could not understand", check that your microphone is set as the default input device in your system settings.

Step 6

Try It Yourself - Challenges

Add "open instagram" and "open twitter" commands
Add a "tell me a joke" command with a few hardcoded jokes
Add a "search for ..." command that searches Google for whatever the user says
Combine with the Weather App tutorial to add a "what's the weather in [city]" command

Requirements