Intermediate · 5h 23m · 22 lessons

AI for Developers

See how LLMs really work, then use AI in your daily dev work without trusting it blindly.

What you'll learn

✦Explain in plain words how AI, ML and LLMs differ, and say which one a problem needs
✦Explain how an LLM turns a prompt into text: tokens, next-token prediction, training and the context window
✦Use AI coding assistants to write, explain, refactor, test and debug code, reviewing every change before you keep it
✦Write prompts with enough context (goal, constraints, code, error) to get usable output, and improve it with focused follow-ups
✦Spot hallucinations, decide what to verify before you trust AI output, and recognise when not to use AI at all
✦Apply responsible AI habits: keep secrets and personal data out of prompts, check for bias and licence risk, and budget token costs
✦Call an LLM API safely from your own code, get structured JSON back, and ground answers in your own data with RAG
✦Evaluate an AI feature with an eval set, add guardrails around it, and explain how agents use tools

Part of these career paths

Syllabus

Module 1 · 1h 5m

AI, ML & LLMs — the landscape Free

What you'll be able to do

Can explain how AI, ML and LLMs nest inside each other, and pick the right one for a problem.
Can describe how an LLM turns a prompt into text: tokens, next-token prediction, training and the context window.
Can name what LLMs are reliably good at and where they reliably fail, and match the check to the claim.
Can match how much they review an AI suggestion to what's actually at stake.

Module 2 · 1h 5m

Where AI fits in a developer's day Pro

What you'll be able to do

Can match how much they review completion, chat and agent-mode suggestions to the mode's blast radius.
Can brief AI to explain unfamiliar code and refactor their own, keeping the interface and behaviour unchanged.
Can brief AI for tests and docs, then read the result for a missing business rule.
Can write a precise debugging prompt, and use AI to learn a new API without skipping the practice.

Module 3 · 57m

Working effectively with AI Pro

What you'll be able to do

Can write a prompt with enough context (role, goal, constraints, code, error) to get a useful first answer.
Can iterate with a targeted follow-up instead of rewriting, repeating or accepting a flawed draft.
Can spot the five common shapes of a hallucination and check each against the right source.
Can decide when a task is a good fit for AI, and when to write it themselves instead.

Module 4 · 59m

Responsible & practical AI Pro

What you'll be able to do

Can decide what is safe to paste into an AI tool, and swap secrets and personal data for placeholders first
Can spot bias in AI-written content and add a review step before it reaches users
Can check AI-written code for licence risk before shipping it
Can estimate the monthly token bill of an AI feature and set limits that keep it in check

Module 5 · 41m

Building with AI APIs Pro

What you'll be able to do

Can call an LLM API from a server, with the key in an environment variable, a max_tokens cap and error handling
Can ask for JSON in a fixed shape and validate it before the code uses it
Can explain the retrieve, add to prompt, answer steps of RAG and say why an answer came out wrong

Module 6 · 36m

AI engineering practices Pro

What you'll be able to do

Can build a small eval set and use it to compare two versions of a prompt
Can place guardrails on the input, the output and the actions of an AI feature
Can trace an agent's tool-calling loop and decide which actions need a human to approve them