The Complete AI Practitioner Course

AI course curriculum: 12 parts, 71 modules

The full syllabus of the course: how AI works, how to use it well at work, how to build with it responsibly, and how to think clearly about where it is going. No coding or technical background needed.

71modules
12parts
18+ hrsguided learning
383knowledge-check questions
311glossary terms

Every module includes a lesson of about a thousand words, an interactive exercise, key takeaways, glossary terms and a knowledge check with an explanation for every answer. The course finishes with a final assessment and a certificate of completion with a verifiable credential ID you can add to LinkedIn.

Enrol for US$299   Read module 1 free

Part 1. Foundations

What AI actually is, how machines learn, and the vocabulary that lets you follow any conversation about it. (6 modules, 88 minutes)

  1. 01What AI is (and isn't)Free to read14 min · True or falseUntangling AI, machine learning, deep learning and generative AI, and why the goalposts keep moving
  2. 02Booms and winters15 min · EstimateSeven decades of promise, disappointment and recovery, and why this cycle looks different
  3. 03How machines learn (no maths)15 min · Put in orderExamples in, patterns found, predictions out, and the one mistake every learner makes
  4. 04Types of learning15 min · Sort into groupsSupervised, unsupervised, reinforcement and self-supervised learning, and where each hides in daily life
  5. 05Data: the raw material15 min · ScenarioWhy quantity, quality and representativeness are three different things, and how data quietly lies
  6. 06The working vocabulary of AI14 min · Match pairsTen terms that let you follow any technical conversation and ask the right question

Part 2. The exponential story

How capability has compounded, how it is measured against humans, and how to think about where it goes next. (5 modules, 76 minutes)

  1. 07Measuring machine intelligence15 min · True or falseWhy benchmarks rise, saturate, and get replaced
  2. 08The scaling era16 min · Sort into groupsHow compute, data and parameters became the levers of progress
  3. 09Cost curves and diffusion14 min · Put in orderHow a fixed capability gets cheaper and spreads through the economy
  4. 10The jagged frontier15 min · ScenarioSuperhuman at some tasks, surprisingly weak at the next one over
  5. 11Forecasting AI without a crystal ball16 min · EstimateTrend lines, bottlenecks, base rates and how to hold calibrated beliefs

Part 3. Under the hood

Neural networks, embeddings, language models, hallucination, retrieval, multimodality and agents, explained for practitioners. (7 modules, 108 minutes)

  1. 12Neural networks and representations15 min · True or falseHow stacks of simple knobs learn features nobody programmed
  2. 13Embeddings: meaning as geometry15 min · Match pairsHow turning words into points in space makes similarity computable
  3. 14Language models16 min · Put in orderTokens, next-token prediction and why fluency is not the same as understanding
  4. 15Why models hallucinateRead an extract15 min · Prompt rewriteThe mechanics of confident error, and how to reduce it
  5. 16Three ways a model "knows"Read an extract16 min · Sort into groupsPre-training, fine-tuning and retrieval, and when to reach for each
  6. 17Multimodal AI15 min · EstimateImages, audio and video share the same underlying ideas as text
  7. 18Agents and tool use16 min · ScenarioFrom answering questions to taking actions, and why that changes the risk

Part 4. The practitioner’s toolkit

Prompting as a discipline, context engineering, verification, iteration and workflow design you can use tomorrow. (8 modules, 121 minutes)

  1. 19Prompting as a disciplineRead an extract15 min · Prompt rewriteBrief a model the way you would brief a capable contractor you have never met
  2. 20Advanced prompting patterns16 min · Match pairsSix reusable patterns, and the honest answer to when each one actually helps
  3. 21Context engineering15 min · Sort into groupsThe model can only work with what is in front of it. Decide what that is.
  4. 22Evaluating outputsRead an extract15 min · True or falseFluent is not the same as right. Build the habits that tell the difference.
  5. 23Iteration loops14 min · ScenarioDraft, critique, refine: how practitioners get to good in fewer, better turns
  6. 24Workflow design16 min · Sort into groupsWhere AI pays off, where it does not, and how to keep a human in the right place
  7. 25AI for analysis and data work15 min · Put in orderAsk sharper questions, describe before you conclude, and check every number
  8. 26AI for writing, research and creative work15 min · ReflectionKeep your voice, check your sources, and be honest about what the machine did

Part 5. Building with AI

System anatomy, build-buy-wait, data readiness, evaluation, trade-offs and honest pilots. (6 modules, 91 minutes)

  1. 27Anatomy of an AI system15 min · Match pairsThe model is one layer of seven, and rarely the one that breaks
  2. 28Build, buy, or wait14 min · ScenarioA decision framework that survives the next price drop
  3. 29Data readiness15 min · Sort into groupsThe unglamorous work that decides whether the glamorous part works
  4. 30Evaluation for AI systems16 min · True or falseWhy 'it looks good' is an opinion, and how to turn it into evidence
  5. 31Cost, latency and quality14 min · Put in orderYou can have two of three, and the skill is choosing which two
  6. 32Running an honest pilot17 min · ReflectionBaselines, controls and the long road from pilot to production

Part 6. Risk, ethics and governance

Bias, privacy, security, reliability, IP, transparency and governance designed to outlast regulatory change. (7 modules, 103 minutes)

  1. 33Bias and fairness15 min · Sort into groupsWhere bias enters a system, why fairness definitions collide, and what you can actually do
  2. 34Privacy and confidentiality14 min · True or falseWhat never goes into a model, how retention really works, and the minimum-necessary habit
  3. 35Security15 min · ScenarioPrompt injection, exfiltration and why model output is untrusted input
  4. 36Reliability and safe deployment16 min · Put in orderFailure modes, graceful degradation, human override and rolling out in stages
  5. 37Intellectual property and attribution14 min · Match pairsTraining data debates, ownership of outputs, disclosure and originality, as principles
  6. 38Transparency, explainability and accountability14 min · ReflectionWhat can be explained, what cannot, and who answers when the system is wrong
  7. 39Governance and regulation: principles15 min · Prompt rewriteRisk-based approaches, policies, audit trails and roles that outlast any particular rulebook

Part 7. AI and the world of work

Tasks versus roles, the T-shaped practitioner, leading adoption, communicating claims and career strategy. (6 modules, 91 minutes)

  1. 40Tasks vs rolesRead an extract15 min · Sort into groupsWhy 'will AI take my job' is the wrong question, and what to ask instead
  2. 41The T-shaped AI practitioner14 min · Match pairsDeep in one domain, fluent across AI, and rich in the skills that appreciate
  3. 42Leading teams through AI adoption16 min · ScenarioChange management for a technology that provokes fear, curiosity and quiet resistance at once
  4. 43Communicating about AI14 min · Prompt rewriteHow to talk to executives, sceptics, customers and regulators without hype or jargon
  5. 44Measuring productivity claims honestly16 min · True or falseWhy most AI productivity numbers are wrong, and how to design measurement you can defend
  6. 45Career strategy in a booming field16 min · Put in orderProof of work, learning loops and community: positioning yourself without chasing every tool

Part 8. Thinking clearly about AI

Hype and doom, cognitive biases, society, alignment and building a personal practice that compounds. (5 modules, 78 minutes)

  1. 46Hype, doom and how to read claims like an expert15 min · ScenarioA repeatable method for separating what a demo shows from what a deployment delivers
  2. 47Cognitive biases in the age of AI14 min · Match pairsThe mental shortcuts that make fluent machines unusually easy to over-trust
  3. 48AI and society16 min · Sort into groupsLabour, information, power and access: the debates and how to hold them fairly
  4. 49The alignment question16 min · True or falseWhy getting a system to want what you want is hard at every scale, from a prompt to a company
  5. 50Building a personal AI practice17 min · ReflectionHabits, a learning loop and a personal eval set that outlast any particular tool

Part 9. AI at work: the productivity playbook

Where the hours go, writing, presentations, business plans, websites without a developer, data analysis, process automation, customer-facing AI, agents, and a 30-day team rollout. (10 modules, 158 minutes)

  1. 51Where the hours go15 min · Sort into groupsMapping a working week into tasks AI can shorten, tasks it can remove, and tasks it should not touch
  2. 52Writing and communication15 min · Prompt rewriteEmails, reports, proposals and meeting notes that sound like you and survive a careful reader
  3. 53Presentations and documents15 min · Put in orderOutline first, generate second: turning a rough brief into a deck that makes one clear case
  4. 54Business plans, strategy and research16 min · ScenarioUsing AI as a sparring partner for plans, market sizing and competitor scans without letting it write your facts
  5. 55Websites, prototypes and simple apps without a developer16 min · True or falseWhat 'describe it and it builds it' can and cannot do, and how to know when to hand over to a professional
  6. 56Data analysis and reporting16 min · ScenarioAsking questions of a spreadsheet without being fooled by the answers
  7. 57Process automation16 min · Put in orderFrom checklists to workflows to agents, and how to choose the first process to automate
  8. 58Customer-facing AI15 min · Match pairsService assistants, recommendations and personalised content: what good looks like and how to measure it
  9. 59Agents in the business17 min · Sort into groupsDelegating multi-step work to software that chooses its own next step, without losing control of it
  10. 60The 30-day team rolloutRead an extract17 min · ReflectionPick, pilot, measure, scale: a week-by-week plan for getting a team using AI well

Part 10. What AI runs on

Tokens, context and cost for business; inside a data centre; the AI supply chain and what it means for pricing and dependency. (3 modules, 46 minutes)

  1. 61Tokens, context and cost for business15 min · EstimateWords in, words out, both cost: how AI usage is metered and how to budget for it
  2. 62Inside a data centre16 min · Match pairsServers, racks, accelerators, megawatts and cooling: the physical layer a business buyer should understand
  3. 63The AI supply chain15 min · True or falseWho designs, makes, hosts, trains and sells, who pays whom, and where the dependency risks sit

Part 11. Case studies: what happened when companies used AI

Recommendation engines, logistics wins, customer service successes and failures, models that met the market, fairness and trust failures, confidentiality and credibility, and the pattern behind them all. (7 modules, 111 minutes)

  1. 64Recommendation engines that built businesses16 min · Put in orderNetflix, Amazon and Spotify, and the long, unglamorous work behind a data flywheel
  2. 65Operations and logistics wins16 min · Match pairsRoutes, robots, engines and weeds: where AI paid off because the problem was narrow and the feedback was fast
  3. 66Customer service, the full picture17 min · ScenarioKlarna, Air Canada, DPD, a Chevrolet dealer and McDonald's: what went right, what went wrong, and who paid
  4. 67When the model met the market17 min · True or falseZillow, IBM Watson and Google Flu Trends: distribution shift, overconfidence and the cost of being wrong at scale
  5. 68Bias, fairness and public trust15 min · ScenarioFour automated decisions that harmed people, and what each teaches about proxies and appeals
  6. 69Confidentiality, content and credibility15 min · True or falseFive incidents where trusting an assistant cost a company its secrets, its accuracy or its standing
  7. 70The pattern behind the cases15 min · Sort into groupsWhat the successes shared, what the failures shared, and how to write your own case study

Part 12. Continuing your learning

Recommended courses and learning paths by goal, how to evaluate any course, and a personal learning plan. (1 modules, 16 minutes)

  1. 71Where to go next16 min · ReflectionWell-known courses and learning paths, grouped by goal, with a way to judge any course and plan your own year

Learn the whole picture

The Complete AI Practitioner Course has 71 modules across 12 parts, about 18 hours of guided learning, an interactive exercise in every module, a final assessment, and a certificate with a LinkedIn credential. No code needed. US$299, paid once.

Enrol nowSee the curriculum