Eight courses to start.

AI Foundations

Thinking Clearly About AI

Most people meet AI as a chat box and decide what it is from there. This course starts somewhere more useful: the difference between asking a question and handing over an outcome. You'll learn what an agent actually does, how to brief one the way you'd brief a capable colleague, and why something that sounds human isn't. You'll treat AI as a partner you learn with rather than a vending machine. You'll also pick up the handful of habits that keep it safe: the least access it needs, approval before anything risky, and a record of what happened. Everything else in the Academy builds on this.

Who it’s for

Everyone, before anything else.

You’ll come away with

A clear picture of what AI and agents are and aren't, and the habits to brief and check them safely.

6 lessons

1. 01

   From answers to outcomes

   Learn what changes when you stop asking AI questions and start handing it outcomes, and how an agent plans, acts and reports back like a project manager.

2. 02

   Briefing an agent

   Brief an agent the way you'd brief a capable colleague: objective, constraints, preferences, deal breakers and outputs. Then use a nine-step blueprint to hand over any task, with checks along the way and a clear point to stop.

3. 03

   Mirrors, not minds

   Why AI sounds human but isn't, how to stop going round in circles with it, and where plain, predictable rules should do the work instead.

4. 04

   Onboard it like a colleague

   Give AI the context, feedback and patience you'd give a new starter, and judge it on how fast it learns, how quickly it works, the quality of its work and how well it adapts.

5. 05

   Learning with AI

   Use AI as a learning partner, not a vending machine, so you understand the answer as well as get it. Simple prompts like "walk me through that" turn nerves into fluency.

6. 06

   Safe by method

   The new risks when AI acts on your behalf, from convincing fakes to hidden instructions, and ten simple rules for using it safely at work. AI gives you a draft, not a decision.

From Using to Building

Personal Software on Demand

A day's training gets most people a ten to fifteen per cent lift, and then they stop. The bigger gain is on the other side of a wall: building with AI, not just using it. This course gets you over that wall. You'll learn to spot the small problems that were never worth solving before. You'll learn to describe the moment you're trying to fix rather than the technology, and to turn it into a working tool in an hour or two. Along the way you'll see eighteen real tools built this way, from a family budget tracker fed by bank texts to a currency converter built in a lift. You'll finish with one of your own, deployed and in use.

Who it’s for

Non-technical people ready for the second lift.

You’ll come away with

A working tool you built for your own problem, online and in use.

7 lessons and a capstone

1. 01

   The two lifts

   Why using AI gives a quick 10–15% boost and then stalls, and how building with it gets you the second, bigger lift.

2. 02

   Personalised Software on Demand

   Cheap building, cheap hosting and an audience of one mean software can now be made just for you. See how seventeen tools went from idea to everyday use in 34 hours.

3. 03

   Small things, at all

   Find the small, everyday annoyances that were never worth fixing before, and see why they're now worth an hour of your morning.

4. 04

   Spec the moment, not the technology

   Describe the moment you want to fix, not how to build it, and have the AI repeat its understanding back before it starts. Clear input, better tools.

5. 05

   The build method

   A step-by-step method from first idea to something other people use, paced over four sessions: prototype, connect real data, put it online, then test and share it.

6. 06

   Find the data you're ignoring

   Spot the useful data already landing on your phone, like bank texts, and turn it into a working tool, with a security check before you share it.

7. 07

   What makes a tool stick

   What makes a tool you keep using: no friction, works offline, respects your attention, and helps you learn as well as get the job done.

Represent First

Prototyping and New Ventures

Ideas stall when they stay abstract, because a concept is hard to buy. This course is about getting an idea out of your head and into a form other people can see, question and act on, before anyone builds it properly. You'll use AI to sharpen a one-line idea into a clear proposition, rehearse the objections, and put up a landing page in an afternoon. You'll learn why high-fidelity prototypes are now as cheap as sketches, and how separate roles are merging into hybrid builders. You'll also see how one person can reach into marketing or sales without retraining. The question changes from "when will someone build this?" to "what can I show today?"

Who it’s for

Founders, product people and intrapreneurs.

You’ll come away with

An idea turned into something people can see, test and buy.

5 lessons

1. 01

   Represent before you build

   People can't buy a concept. Learn a six-step method that takes an idea from one line to something people can see and react to, using the launch of this Academy, done in about two hours, as the worked example.

2. 02

   Speed is the resolution

   High-fidelity prototypes now cost about the same as sketches. Learn to make them fast, and to get feedback on the idea rather than the button colours.

3. 03

   The great skill renegotiation

   As designer, analyst and developer roles merge into hybrid builders, a small team can test five ideas in the time one used to take.

4. 04

   Reach into adjacent roles

   How one person runs a full content operation for under $60 a month, and how to point the skills you already have at a new job.

5. 05

   Spark, don't replace

   Use AI to generate and test ideas from what you've already written, while the writing itself stays yours.

Redesigning Work from First Principles

Most AI adoption adds a tool to the gaps in an old process and calls it transformation. This course starts from the other end: assume most tasks in a role can be automated, then ask what the job is now. You'll learn to choose between training the person, fixing the process, and removing the task altogether. You'll design workflows that run AI-to-AI by default, with people only at the gates that earn their place. And you'll test a redesigned team against the existing one before switching over. It finishes with the cost side. Tokens are the new hours, and the cheapest work is the work you've turned into a script.

Who it’s for

Managers, operations and process owners.

You’ll come away with

One real workflow, redesigned and tested against the old way.

6 lessons

1. 01

   Tasks, not jobs

   AI will automate most of the tasks in a job, not most of the jobs. Learn to separate the human work from the work AI can run, and six moves to future-proof your role.

2. 02

   Fix the person, fix the process, or remove the task

   Three questions to decide whether to train people, redesign the process or drop the task altogether, and a simple way to test before you scale.

3. 03

   AI-to-AI by default

   Design workflows where AI hands work to AI and people step in only where they're truly needed: decisions you can't undo, accountability, real ambiguity or taste.

4. 04

   Contracts, not job descriptions

   Define work by what goes in and what comes out, build an AI-first version of a team alongside the current one, and switch over only when it's proven.

5. 05

   The uncomfortable maths

   If everyone gets faster, you can cut, grow or give time back. Why to choose openly, and why people fear silence more than they fear AI.

6. 06

   The token economy

   AI is paid for by usage, so tokens are the new hours. Learn to send each task to the right model, turn repeat work into scripts, and keep costs down.

Leading a Team of Agents

Becoming a HALO

A HALO, or Human-Agent Led Operator, is one person backed by a team of agents and accountable for a whole outcome. This course is about becoming one. You'll learn to run several streams of agent work at once without losing track of the one that has stalled. You'll set the gates where your judgement is needed and stay out of the rest. And you'll direct agents with the precision they need: a clear definition of done, honest feedback, and standards that don't move. It closes on the question every HALO eventually faces: when the capability lives in your agents, who owns it?

Who it’s for

Practitioners who will run work through agents.

You’ll come away with

Several streams of agent work running at once, each with clear checkpoints and in plain view.

6 lessons

1. 01

   The HALO

   What a Human-Agent Led Operator is: one person with a team of agents, accountable for a whole outcome. You don't have direct reports; you have direct outputs.

2. 02

   Octopusing

   Run several streams of work at once and notice quickly when one needs you. The skill is switching cleanly between them, and some people in your organisation already have it.

3. 03

   Design for visibility

   Set up one view of everything your agents are doing, read-only by default, so nothing stalls unseen.

4. 04

   Managing agents is not managing people

   Agents need precision: a clear definition of done, honest feedback and standards that don't slip.

5. 05

   Delegated intent

   Step in only when something is off, and know when to restart, abandon or escalate a piece of work.

6. 06

   Who owns the capability

   When your agents hold the know-how, who owns it? What that means for your role, your employer and how you're paid.

Leading AI Adoption

The technology is ready; most organisations aren't, and that is a leadership problem. This course is for the people responsible for adoption. It explains why adoption so often surges after the first push and then fragments, and what stops that. The answer is leaders who go first, a clear reason for the change, honesty about what happens to jobs, and a culture built through small, visible wins rather than memos. You'll look at what HR becomes in an organisation built around systems rather than people, and what happens to managers. You'll see how hiring changes when skills can be learned in minutes. And you'll see how a Build Clinic turns a training day into lasting capability.

Who it’s for

Executives, HR and L&D.

You’ll come away with

An adoption plan that keeps going after the first push fades.

8 lessons

1. 01

   A leadership problem, not a technology one

   Why adoption surges after the first push and then fragments, and how a clear purpose and honesty about jobs keep it moving.

2. 02

   Leaders go first

   An organisation only goes as deep as its leaders do. Why the top team has to use AI themselves, and who should own it.

3. 03

   Culture through small wins

   Get past the middle-management bottleneck with small, visible wins and a twice-weekly fast lane for decisions.

4. 04

   Built for systems, not people

   What changes when work runs through systems as well as people: HR as the architect of culture, and team leaders becoming systems coaches.

5. 05

   Three managers, three fates

   Builders, enablers and title-holders: which managers thrive, which don't, and why to reward depth of skill over the size of a team.

6. 06

   Hire for learning

   When skills can be picked up in minutes, hire for how people learn. What to look for, the return of apprenticeships, and how to avoid age bias.

7. 07

   Make the second lift stick

   Turn a single training day into lasting capability with Build Clinics: half-day sessions, one supervisor to five builders, on a regular rhythm.

8. 08

   Internal builders

   Why you need your own builders again, who should own each internal tool, and why the model is not the product.

Reinventing the Business

AI doesn't just change how work gets done; it changes what a business sells. This course follows that change through: from tools to outcomes, from generic software to products shaped around one customer, from billing hours to pricing results. You'll see how a single founder can run departments made of agents, and why an organisation that never bought the big central system may now move fastest. You'll learn what it means to model how a business actually works rather than search its documents. It ends with the questions of ownership that follow when the process, not the finished piece, becomes the product.

Who it’s for

Founders, the C-suite and strategy.

You’ll come away with

One function to own end to end, and pricing that rewards results rather than hours.

8 lessons

1. 01

   The five stages of AI in business

   Find where your business sits, from "sprinkle some AI in" to "AI just is", and why selling the outcome beats selling the tool.

2. 02

   Generic software is over

   Why one-size-fits-all software is giving way to tools shaped around one customer at the moment they need them, and what that means for what you sell.

3. 03

   The end of time-based value

   AI is dissolving the billable hour. Four ways to price instead: outcomes, judgement, access and services that compound over time.

4. 04

   The solo founder

   How one person can run departments made of agents, bring in specialists for a mission, and grow without growing headcount.

5. 05

   The leapfrog

   Why organisations that never built the big central system may now move fastest, and three routes to choose between.

6. 06

   Model the business

   Move from searching documents to modelling how your business actually works, so it can ask its people for what it doesn't yet know.

7. 07

   Personalisation as architecture

   Build around what customers want, not just what they click. Audit the customer journey and keep privacy built in from the start.

8. 08

   Protect the verb

   When the process becomes the product, protect how you work: where outputs came from, digital twins of roles, and rules for owning a model.

Building with AI for Young People

Young people don't carry the hesitation about AI that many adults do: they expect to describe something and have it happen. This short course is for the parents and teachers who want to turn that into building. It uses real tools rather than children's apps. It starts from what the young person cares about and what annoys them. Then it works through a guided plan over four sessions, from a first prototype to something live on their phone. The aim is simple: "I had an idea" followed by "so I built it".

Who it’s for

Parents, teachers and young builders.

You’ll come away with

A real thing your young person has built and shipped.

3 lessons

1. 01

   Real tools, not toy ones

   Children's AI apps hit a wall fast. Set up real tools safely, because the barrier is configuration, not intelligence.

2. 02

   Passion plus frustration

   Start from what they love and what annoys them, generate twenty ideas, and narrow down to something they can finish this week.

3. 03

   The Four-Session Plan, guided

   Four guided sessions from a first prototype to something live on their phone, including what to fake for now and what to make real later.