Creating a Chatbot for a Restaurant

pip install openai panel

import os
import openai
import panel as pn

# Initialize Panel extension
pn.extension()

# Configure your OpenAI API key from environment variable
openai.api_key = os.getenv("OPENAI_API_KEY")
# Or hardcode your key (not recommended for production):
# openai.api_key = "YOUR_API_KEY"

def get_chat_completion(messages):
    """
    Sends a list of message dicts to OpenAI and returns the assistant response.
    """
    response = openai.ChatCompletion.create(
        model="gpt-3.5-turbo",
        messages=messages
    )
    return response.choices[0].message.content

def collect_messages(_):
    # 1. Read and clear user input
    prompt = inp.value
    inp.value = ''

    # 2. Append user message and get reply
    context.append({'role': 'user', 'content': prompt})
    response = get_chat_completion(context)
    context.append({'role': 'assistant', 'content': response})

    # 3. Update the chat panels
    panels.append(pn.Row("**User:**", pn.pane.Markdown(prompt, width=600)))
    panels.append(
        pn.Row(
            "**Assistant:**",
            pn.pane.Markdown(response, width=600),
            styles={'background-color': '#F6F6F6'}
        )
    )
    return pn.Column(*panels)

bot_instructions = """
You are BurgerBot, an automated assistant for Burger Bliss. 
1. Greet the customer.
2. Collect the full order.
3. Ask if it’s pickup or delivery.
4. Summarize the order and confirm.
5. If delivery, request the address.
6. Collect payment.
Respond in a friendly, short conversational style.
"""

---

## 5. Building the UI with Panel

Initialize the conversation context and Panel widgets. Bind the “Chat” button to our `collect_messages` function.

```python
# Initialize state
panels = []
context = [{'role': 'system', 'content': bot_instructions}]

# Create input widget and button
inp = pn.widgets.TextInput(value="", placeholder="Type your message here...")
button_conversation = pn.widgets.Button(name="Chat", button_type="primary")

# Bind callback
interactive_conversation = pn.bind(collect_messages, button_conversation)

# Assemble the dashboard layout
dashboard = pn.Column(
    pn.pane.Markdown("## 🍔 Burger Bliss Chatbot"),
    inp,
    pn.Row(button_conversation),
    pn.panel(interactive_conversation, loading_indicator=True, height=400),
)

# Display the UI
dashboard.show()
```text

---

## 6. Example Conversation & Inspecting Context

After launching the dashboard, try these sample exchanges:

1. “I want to order a chicken burger.”  
2. “Spicy Crispy Chicken Burger.”  
3. “Can I have the bill please?”  
4. “I am done.”  
5. “What’s the total?”  
6. “Delivery”

To inspect the stored message history:

```python
# In a separate cell, print the context
print(context)
```text

You’ll see a list of dictionaries:

```python
[
  {'role': 'system',    'content': '...bot instructions...'},
  {'role': 'assistant', 'content': 'Hi! Welcome to Burger Bliss!'},
  {'role': 'user',      'content': 'I want to order a chicken burger.'},
  {'role': 'assistant', 'content': 'Great choice! Which chicken burger would you like?'},
  ...
]