“`html

How to Build Your First AI Assistant with Python – A Practical Step-by-Step Tutorial

1. Setting Up Your Development Environment

• Install Python 3.9+ and create a dedicated virtual environment (e.g., python -m venv ai_env).
• Sign up for an OpenAI account, create an API key, and install the openai library via pip.
• Verify the setup by running a minimal “Hello World” completion call using the gpt-3.5-turbo model.

2. Understanding the Core Concepts: Prompts and Completions

• Learn the role of system, user, and assistant messages in the chat completion format.
• Craft clear and specific prompts to guide the model toward the desired output.
• Experiment with the temperature (0–2) and max_tokens parameters to control creativity and response length.

3. Writing the Basic Chat Loop

• Create a Python list (e.g., messages = []) to store the full conversation history.
• Implement a while True loop that accepts user input and appends it to the messages list.
• Call the OpenAI API with the updated messages list and print the assistant’s response in real time.

4. Adding Personality and Constraints with System Messages

• Prepend a system message (e.g., “You are a friendly coding tutor”) to set the assistant’s behavior and tone.
• Use system instructions to enforce constraints – for example, “Only answer in one sentence” or “Never mention competitors.”
• Test multiple system prompts side by side to observe how the same user input yields different responses.

5. Handling Errors and API Rate Limits

• Wrap every API call inside a try-except block to catch openai.error.APIError, RateLimitError, and AuthenticationError.
• Implement exponential backoff logic (using time.sleep()) when a rate limit is hit.
• Log errors to a file and provide user-friendly fallback messages instead of crashing.

6. Enhancing with Simple Memory and Context

• Maintain contextual conversation history by keeping the messages list across loop iterations.
• Implement a token counting function to truncate the oldest messages when the total exceeds a cost threshold (e.g., 3000 tokens).
• Optionally save the conversation to a local JSON file so users can review past chats.

7. Deploying Your Assistant as