So, you’re a builder, huh? Someone who loves getting into the code, having total say, and turning your creations into things people can buy? Cool. It’s not some crazy hard thing to do. Basically, you just need a brain and a body to get started.

For a local business, we don’t need a massive neural network. We just need to map keywords to responses. A simple Python dictionary, or even better, a JSON file, is the perfect “brain.”

Create a file named (intents.json):

"greetings": {
    "keywords": ["hi", "hello", "hey", "good morning"],
    "response": "Hello! We are open 9 AM to 5 PM. How can I help?"
  },
  "hours": {
    "keywords": ["hours", "open", "when", "close"],
    "response": "Our hours are 9 AM to 5 PM, Monday to Friday."
  },
  "location": {
    "keywords": ["where", "address", "find you", "location"],
    "response": "You can find us at 123 Main Street, Downtown."
  },
  "default": {
    "response": "I'm sorry, I can only help with hours and location. Please call (555) 123-4567 for more help."
  }
}

Create a file named (bot_logic.py) and paste this code in:

import json
import string

def load_intents(filename="intents.json"):
    """
    Loads the intent file and returns the data.
    """
    try:
        with open(filename, 'r') as file:
            intents = json.load(file)
        return intents
    except FileNotFoundError:
        print(f"Error: The file {filename} was not found.")
        return None
    except json.JSONDecodeError:
        print(f"Error: Could not decode the JSON file {filename}.")
        return None

def clean_message(message):
    """
    Cleans the user message:
    1. Converts to lowercase.
    2. Removes all punctuation.
    """
    message = message.lower()
    # str.maketrans creates a translation table to remove punctuation
    message = message.translate(str.maketrans('', '', string.punctuation))
    return message

def get_bot_response(user_message, intents):
    """
    Finds and returns the correct bot response based on keywords.
    """
    if not intents:
        return "Sorry, the bot is not configured."
        
    cleaned_message = clean_message(user_message)
    
    # Split the user message into individual words
    message_words = set(cleaned_message.split())

    # Check for a match in keywords
    for intent, data in intents.items():
        if intent == "default":  # Skip the default case for now
            continue
            
        # Check if any keyword matches any word in the user's message
        for keyword in data["keywords"]:
            if keyword in message_words:
                return data["response"]
    
    # If no keyword was found, return the default response
    return intents["default"]["response"]

# --- THIS IS HOW YOU TEST YOUR BOT ---
# This part only runs when you execute 'python bot_logic.py' directly
if __name__ == "__main__":
    intent_data = load_intents()
    
    if intent_data:
        print("Bot is ready! (Type 'quit' to exit)")
        
        while True:
            # Get input from the user in the terminal
            user_input = input("You: ")
            
            if user_input.lower() == 'quit':
                break
            
            # Get the bot's response
            response = get_bot_response(user_input, intent_data)
            
            # Print the response
            print(f"Bot: {response}")

How to Use This (The 100% Solution)
1. Save both files (intents.json and bot_logic.py) in the same folder.
2. Open your terminal (command prompt) and navigate to that folder.
3. To test it, run this command: python bot_logic.py
4. Your bot will say “Bot is ready!” right in your terminal.
5. You can type “when are you open?” and test it.

That’s it. You now have a complete, functional “brain.” Your bot_logic.py file is fully operational. You don’t need to change this code again. Your next step will be to call this get_bot_response function from within your Flask API.

You’ve built a powerful “brain,” but right now, it only lives in your terminal. To make it a real product you can sell, it needs a “body” a way to connect to the internet, receive requests, and send back answers. This is where a Flask API comes in. Flask is a lightweight Python web framework that is perfect for this job. It lets us create a simple web server in just a few lines of code.

File 3: The “API” (app.py)
1. Import your bot_logic (the brain).
2. Start a web server.
3. Create a single API endpoint (a “URL”) named /chat.
4. Listen for incoming JSON messages at that URL.
5. Pass the message to your get_bot_response function.
6. Send the function’s answer back as JSON.

Create a new file named app.py in the same folder as bot_logoic.py and intents.json.

from flask import Flask, request, jsonify
# You would import your bot_logic.py here
# from bot_logic import get_bot_response 

app = Flask(__name__)

@app.route("/chat", methods=["POST"])
def chat():
  # Get the message from the user
  user_message = request.json.get("message")

  # 1. Get response from your bot_logic "brain"
  # bot_response = get_bot_response(user_message) 
  
  # For this example, let's just send a test response
  bot_response = f"You said: {user_message}" 

  return jsonify({"response": bot_response})

if __name__ == "__main__":
  app.run(debug=True)

Install Dependencies: Before running, they must install Flask and Flask-CORS. Have them open their terminal and run:

pip install Flask flask-cors

Check Files: Ensure all three files (intents.json, bot_logic.py, and app.py) are in the same folder.
Run the Server: In the terminal, run this command:

python app.py

The Result: The terminal will come to life and show something like this:

* Serving Flask app 'app'
* Running on http://127.0.0.1:5000
* Running on http://192.168.1.10:5000 (Press CTRL+C to quit)
* Restarting with stat

Pro & Con for Your Client (The Sales Pitch)
1. Pro: This is your 100% custom-built asset. The biggest selling point is “No Monthly SaaS Fees.” The client doesn’t have to pay a subscription to Tidio or Chatbase. You can charge a one-time setup fee (e.g., $500) and a small monthly “hosting & maintenance” fee (e.g., $50/month) that goes directly to you.
2. Con: The “time-to-market” is slower. You can’t add complex features like live bookings or inventory checks in one day. This solution is perfect for FAQ & info, but it’s complex to scale without more development time.

If you want to master the AI skills behind this project, dive deeper with our [full-stack roadmap to learn AI development with Python]