Building a Product-Query Chatbot with OpenAI & Python

If you’re running an e-commerce store, you’re familiar with the countless questions customers have about products. Today, we’ll build a chatbot that can provide product details based on a user’s queries. We’ll be using Python along with the OpenAI API.

1. Setting up the Environment

Before diving into the code, ensure you have the prerequisites:

• Install Python from python.org.
• Set up a virtual environment. This isn’t mandatory but can be helpful:

python -m venv myenv
source myenv/bin/activate  # On Windows use `myenv\Scripts\activate`

Install the necessary packages:

pip install openai

2. Preparing the Product Data

Our bot will fetch product details from a JSON file. Let’s structure our data:

[
    {
        "product": "Smart TV",
        "details": "42-inch 4K UHD",
        "price": "$500",
        "delivery_time": "5-7 days",
        "estimated_delivery": "7th Oct",
        "location": "New York",
        "buy_link": "http://example.com/smarttv"
    },
    // ... add more products as you like
]

3. Crafting the Chatbot Logic

Here’s a breakdown of the Python code for our chatbot:

a) Initialize OpenAI and Load the Data

import openai
import json

# Initialize OpenAI
openai.api_key = 'YOUR_API_KEY'

with open('product_data.json', 'r') as file:
    products = json.load(file)

b) Querying OpenAI and Getting Product Details

We define a function to get details of a specified product:

def get_product_detail(product_name):
    for product in products:
        if product_name.lower() in product["product"].lower():
            return product
    return None

c) Initiating the Chat

The chat function will guide the interaction:

def chat():
    system_prompt = "Welcome! I'm here to help you find products. What product are you interested in?"
    print(system_prompt)
    
    product_name = input("You: ")
    product_detail = get_product_detail(product_name)
    
    if not product_detail:
        print("Sorry, I couldn't find that product.")
        return

    details_to_ask = ["details", "price", "delivery_time", "estimated_delivery", "location"]
    for detail in details_to_ask:
        question = f"Would you like to know the {detail.replace('_', ' ')} of {product_name}?"
        print(question)
        
        user_response = input("You: ")
        if "yes" in user_response.lower():
            print(f"The {detail.replace('_', ' ')} of {product_name} is {product_detail[detail]}.")

    print(f"Would you like to buy the {product_name}? Here's the link: {product_detail['buy_link']}")

chat()

4. Making the Code Production-Ready

Consider these steps for a robust chatbot:

1. Diversify your product JSON data.
2. Handle errors for product unavailability or API issues.
3. Add user confirmation before showing purchase links.
4. For a UI-based experience, integrate with Flask or Django.
5. Ensure the OpenAI API key is stored securely.
6. Implement logging for troubleshooting.
7. Conduct thorough testing with diverse inputs.

In Conclusion

The above implementation provides a basic yet comprehensive chatbot for querying product details and guiding potential customers. Remember to conduct robust testing before taking the chatbot live.

Happy coding and selling!