Your whole team moves, not just the early adopters.
AI adoption tends to widen the gap between your strongest and weakest engineers. This course raises the floor. The middle group catches up, standards get shared, and the gap narrows.
Launching soon
Most engineers are using AI. Very few are using it well.
Not because of the tools. Because of how they think about the work.
There's no shortage of AI content. What's missing is a framework built by someone who's done it with real teams, that respects your experience and gets to the point.
This course teaches the mental models and workflows that make the difference. Less about which tool to use this week, more about how to direct AI well, review what it produces, and stay accountable for what ships.
Be the first to know when it launches, along with any early access pricing.
Most engineers are using AI. Very few are using it well. The gap isn't tools. It's judgment, workflow, and mindset. This course is the difference between the two.
Most engineers are using AI. Very few are using it well.
The gap isn't tools. It's judgment, workflow, and mindset.
This course is the difference between the two.
If you lead an engineering team, this is for you.
Most teams have the same distribution.
A few engineers pushing AI hard. A few who aren't interested. And a large middle group who are curious, capable, and not yet effective.
That middle group is your biggest lever. They already have the judgment to direct AI well. They just haven't built the workflows yet.
This course gives them a shared foundation: how to scope work for AI, write a brief that gets good output, review what comes back, and take accountability for what ships. Not a tool tutorial. A way of working.
Less fragmentation, more shared language.
When your team has the same vocabulary and framework, you spend less time re-explaining decisions and more time building. Shared context compounds.
Built on the three problems every team has to solve.
Context, guardrails, and orchestration. The course covers each one practically, so engineers leave knowing not just what to do but why.
If you're the engineer, this is for you too.
Not catching up. Getting ahead.
A lot of engineers are quietly worried that AI makes their work less interesting. That they become reviewers of generated code rather than engineers who build things. That the craft disappears.
Done badly, that worry is justified. Done well, the opposite is true.
This course is about doing it well. You'll learn how to direct AI clearly, decompose work so it gets good output, review what comes back critically, and stay accountable for what ships. These are skills that make you more capable, not less relevant.
And they make the work more interesting. More autonomy. More time on the problems worth solving. Less time on the parts you never enjoyed anyway.
You'll know how to get consistently good output from AI.
Not occasionally, when the prompt happens to work. Reliably, because you know how to scope the task, write the brief, and review what comes back.
You'll stay sharp, not become dependent.
The real risk with AI isn't that it takes your job. It's that you stop thinking and let it drive. This course is built around keeping you in the loop — directing the work, reviewing the output, and owning what ships. Your judgment stays central. The AI does the lifting.
You'll have a framework that survives tool changes.
The tools will keep changing. The way you think about decomposing work, directing AI, and staying accountable for what ships will not.
You'll feel more in control, not less.
A lot of engineers feel like AI is something that happens to them. This course is about being the one who directs it.
What's in the course
What it means to work with AI by default. Not a new tool in your stack. A different way of thinking about the work. Covers the two failure modes most engineers fall into, why "what can it do?" is the wrong question, and what the shift from writing code to directing it actually looks like in practice.
These concepts are foundational in the way that HTTP is foundational. You learn them once and they travel. Tokens, context windows, models, modes, and MCP are not implementation details of a specific tool. They are the structure beneath every tool you will ever use. This section builds the precise vocabulary the rest of the course depends on, and explains why some of these ideas, context windows especially, have real practical consequences for how you work.
The core of the course. Working with AI is less like typing and more like managing a capable junior engineer. Covers onboarding your AI to your codebase, breaking down work, writing briefs that get good output, setting defaults, deciding when to stay in the loop and when to step back, building in escalation, using your existing guardrails, reviewing output critically, and taking accountability for what ships.
Individual fluency does not automatically become team capability. This section covers the three problems every team needs to solve: getting AI up to speed on your context, making sure your guardrails are strong enough to support faster output, and finding your team's right level of autonomy. Plus how to share what works, maintain standards without mandating tools, and measure progress in ways that actually matter.
What is getting better, what is staying stable, and how to evaluate new tools without chasing hype. Covers the direction of travel in AI-assisted development and what it means for the role of the senior engineer as judgment, accountability, and direction become the work.
Sam Jarman
I'm an engineering manager who thinks most AI adoption advice misses the point. It focuses on tools and tutorials when the real challenge is judgment. Knowing how to direct AI well, review what it produces, and stay accountable for what ships.
I've spent years leading engineering teams through this shift, writing about it at samjarman.co.nz, and figuring out what actually works in production codebases with real engineers under real deadlines. Not in demos. Not in toy projects.
This course is the distillation of that. Built for the engineers I wish I'd had a sharper way to help sooner.
Join the waitlist.
The course isn't live yet. Leave your email and we'll reach out when it launches, along with any early access pricing.
We'll use your email for launch updates, early access details, and occasional course news. Read the privacy policy.