AI Assistant Martian Using Python

Martian is a voice-controlled AI assistant where all commands and responses are stored in a CSV dataset. This approach teaches readers how to structure data, access it using Python, and make the assistant data-driven - a core concept in machine learning projects. Instead of hard-coding responses, Martian dynamically reads from the CSV to respond to user queries.

Python Code

# Martian AI - Python Example
# Author: Neelkanth

import pyttsx3
import speech_recognition as sr
import datetime
import wikipedia
import webbrowser
import pandas as pd
import os

# Load CSV Dataset

DATA_FILE = "martian_dataset.csv"

if not os.path.exists(DATA_FILE):
    data = {
        "command": ["name of your creator", "how far is mars from earth", "how are you",
                    "who are you", "open youtube", "open google", "open whatsapp", "the time",
                    "wikipedia python"],
        "response": ["Neelkanth", "On average, Mars is 225 million kilometers away from Earth.",
                     "I am fine", "I am Martian, your personal assistant.",
                     "Opening YouTube", "Opening Google", "Opening WhatsApp",
                     "[Current Time]", "[Wikipedia Summary]"]
    }
    df = pd.DataFrame(data)
    df.to_csv(DATA_FILE, index=False)
else:
    df = pd.read_csv(DATA_FILE)

#Text-to-Speech

engine = pyttsx3.init('sapi5')
engine.setProperty('voice', engine.getProperty('voices')[0].id)

def speak(text):
    engine.say(text)
    engine.runAndWait()

# Greetings

def wish_me():
    hour = datetime.datetime.now().hour
    if hour < 12:
        speak("Good morning Sir!")
    elif hour < 17:
        speak("Good afternoon Sir!")
    else:
        speak("Good evening! Have a great day!")

# Voice Input

def take_command():
    r = sr.Recognizer()
    with sr.Microphone() as source:
        print("Listening...")
        r.pause_threshold = 1
        audio = r.listen(source)
    try:
        query = r.recognize_google(audio, language='en-in')
        print(f"User said: {query}")
    except Exception:
        print("Could not understand audio, please repeat.")
        return "none"
    return query.lower()

# CSV Lookup

def get_response(query):
    row = df[df['command'].str.lower() == query.lower()]
    if not row.empty:
        response = row['response'].values[0]
        if response == "[Current Time]":
            response = datetime.datetime.now().strftime("%H:%M:%S")
        elif response == "[Wikipedia Summary]":
            try:
                topic = query.replace("wikipedia", "")
                response = wikipedia.summary(topic, sentences=2)
            except:
                response = "No results found on Wikipedia."
        if "open youtube" in query:
            webbrowser.open("https://youtube.com")
        elif "open google" in query:
            webbrowser.open("https://www.google.com")
        elif "open whatsapp" in query:
            webbrowser.open("https://web.whatsapp.com")
        return response
    elif query in ["quit", "exit", "close"]:
        return "exit"
    else:
        return None

# Main Loop

if __name__ == "__main__":
    wish_me()
    speak("I am Martian, your CSV-driven assistant. How can I help you today?")
    while True:
        query = take_command()
        if query == "none":
            continue
        response = get_response(query)
        if response == "exit":
            speak("Thank you for using Martian. Goodbye!")
            break
        elif response is None:
            speak("I don't know this command yet. Try another one.")
        else:
            speak(response)

CSV Dataset Example

This CSV file stores the commands and their corresponding responses. Users can open this CSV to understand how the AI assistant accesses structured data.

Key Features

CSV-Driven Dataset

All commands and responses are stored in a CSV file, demonstrating how AI systems can be data-driven and easily extensible.

Voice Recognition

Uses the SpeechRecognition library to convert spoken words into text, allowing hands-free interaction.

Task Automation

Recognizes predefined commands to open websites like Google, YouTube, WhatsApp using Python's webbrowser module.

Dynamic Responses

Some responses like current time or Wikipedia summaries are generated dynamically, making Martian flexible.

Workflow

Step 1 - Voice Input

Step 2 - CSV Lookup

Martian searches the CSV dataset for a matching command and retrieves the corresponding response.

Step 3 - Response Execution

Martian speaks the response and, if applicable, opens websites or fetches dynamic data like Wikipedia summaries.

Limitations

No NLP Integration: Commands must match exactly with CSV entries. Martian cannot understand complex natural language.

Future Enhancements

NLP Integration: Use NLP to understand user intent instead of exact command matching.

API Integration: Connect Martian with external APIs for music, weather, or news.

Conclusion

This CSV-driven Martian AI assistant teaches readers how to use structured data to build a real-time, voice-controlled assistant. It demonstrates the basics of AI workflows, dynamic responses, and data-driven design in Python.

AI Assistant Martian Using Python and CSV Dataset

By Neelkanth

Project Overview

Python Code

CSV Dataset Example

Key Features

CSV-Driven Dataset

Voice Recognition

Task Automation

Dynamic Responses

Workflow

Step 1 - Voice Input

Step 2 - CSV Lookup

Step 3 - Response Execution

Limitations

Future Enhancements

Conclusion

AI Assistant Martian Using Python and CSV DatasetBy Neelkanth

Project Overview

Python Code

CSV Dataset Example

Key Features

CSV-Driven Dataset

Voice Recognition

Task Automation

Dynamic Responses

Workflow

Step 1 - Voice Input

Step 2 - CSV Lookup

Step 3 - Response Execution

Limitations

Future Enhancements

Conclusion

AI Assistant Martian Using Python and CSV Dataset

By Neelkanth