LLM API Intro: Plug Model Capabilities into Your App

What you're probably confused about right now

"Isn't knowing how to write prompts enough?"

In a real product, you need code that calls the model automatically -- and handles failures, retries, and logging.

One-line definition

Calling an LLM API is a full engineering workflow: auth, request, parse, error handling, and retry.

Real-life analogy

Ordering food delivery isn't just picking dishes. You also handle payment, waiting, following up, and dealing with wrong orders.

Minimal working example

import os
API_KEY = os.getenv("API_KEY")
if not API_KEY:
    raise ValueError("missing key")

Quick quiz (5 min)

1. Implement ask_llm(prompt).
2. Add timeout and try/except.
3. Retry twice on failure and log each attempt.

Quiz answer guide & grading criteria

• Answer direction: write runnable code that covers the core requirements and edge cases from the prompt.
• Criterion 1 (Correctness): Main flow produces correct results, key branches execute.
• Criterion 2 (Readability): Clear variable names, no excessive nesting.
• Criterion 3 (Robustness): Basic protection against null values, type errors, or unexpected input.

Take-home task

Build a "Q&A interface wrapper" with unified error codes and response format.

Acceptance criteria

You can independently:

• Run a minimal API call end-to-end
• Handle timeout/auth/empty response errors
• Add basic observability logging

Common errors & debugging steps (beginner edition)

• Can't read the error: start from the last line -- find the error type (TypeError, NameError, etc.), then trace back to the line in your code.
• Not sure about a variable's value: throw in a temporary print(var, type(var)) at key points to verify data looks right.
• Changed code but nothing happened: make sure the file is saved, you're running the right file, and your terminal is in the correct venv.

Common misconceptions

• Misconception: if you got a response, it's a success.
• Reality: you need to validate the content and check for stability.