Why this course, and the no-jargon contract

You opened this, which is already more commitment than most people give a new AI course before they bounce, so let's not waste it.

Here's what just happened, in the world. A lot of very smart people in the last eighteen months have tried to do something concrete with AI, whether that's analyzing data, automating a workflow, shipping a small internal tool, or building a research site. Many of them quietly gave up, not because the AI failed them, but because at some point Claude or ChatGPT or Cursor asked them a reasonable-sounding question like do you want this hosted as a serverless function or on a VM?, and they realized they didn't know what the question meant, what the right answer was, or which one would cost them four thousand dollars by Friday.

That moment is what this course is about.

Technical Enough is 21 short lessons of about fifteen minutes each, and they teach the vocabulary and mental model AI tools assume you already have. There is no code, you won't deploy anything, and you won't open a terminal. By Day 21 you'll have two things: a working understanding of how modern software actually gets built and shipped, and a one-page Build Plan for the specific thing you want to build.

That is the whole pitch. We are not promising you a new job, a startup, or a six-figure side project. We're promising that the next time Claude asks you a perfectly reasonable question, you will not stare at the screen.

Who this is for, and who it isn't

This course is for you if any of these sound familiar:

• You're a product manager and the floor under your role has been shifting and you can feel it.
• You're a domain expert (a scientist, a doctor, a lawyer, a marketer, an ops lead) who tried to build something with AI and kept getting stuck on what should have been the easy parts.
• You're someone who reads every AI essay published this year and still can't quite explain what a "deployment" is to your mother.

This course is not for you if you already know what a Docker container is and what's inside it. We won't teach you the craft of software engineering, because twenty other courses do that and most of them do it well. We're teaching the layer underneath the prompts, once, plainly, in three weeks.

If that is not what you came for, this is your honest exit, feel free to close the tab and there will be no hard feelings.

The no-jargon contract

For everyone still here, the deal.

Every technical term that appears in this course gets translated into plain English the first time it shows up, and then it goes into a glossary that grows with you across all 21 lessons. There is not a single sentence in this course that contains a term we haven't defined yet.

This is the part where most technical courses quietly break their own promise. They define API once, on page fourteen, in a sentence that also contains REST, endpoint, and payload, and now you're three jargon-debts deep before you even finished the definition. We're not going to do that.

When a word like serverless shows up later, it lands like this:

Serverless. A way of renting computer time on the internet where you pay only when your code is actually running, not all the time. The name is a lie, because there are servers, you just don't have to think about them.

Every term, every time, with the small italic note that says: the name is a lie, and here is what it actually is.

The 21-day map

Three weeks, and each one has a job.

Week 1, The Map. Seven lessons that draw the territory: what software actually is (Day 2), where it lives on your laptop versus on the internet (Day 3), the full journey a button-press takes (Day 4), and the things underneath like languages, files, environments, and version control. No tools opened, no code written. You finish Week 1 able to look at any piece of software and say what it is, where it runs, and how it gets there.

Week 2, The Layers. Seven lessons, one per layer of the stack: frontend, backend, APIs, data, hosting, auth, monitoring. Each lesson opens with the exact question an AI tool will ask you about that layer, and ends with your answer to it. Day 12 is the hosting question that broke the scientist in the founding essay, and you'll answer it before lunch.

Week 3, Your Build. Seven lessons where we apply the map. You pick what you want to build (Day 15), you decide whether it's a web app, a mobile app, or actually a script (Day 16), you decide where AI fits and where it explicitly doesn't (Day 17), you sketch your stack, you learn the guardrails experienced builders wish someone had told them (Day 19), and the legal sanity checks nobody mentions until after you've already broken one (Day 20). On Day 21 you sequence your build, submit a one-page Build Plan, and walk out with a finisher's card you can post anywhere.

That is the shape. If you're skimming for the punchline, by Day 21 you'll be able to look at the thing you want to build, name every layer of it, and write the one-sentence version of what to build first.

How to actually do this

A few notes on usage that sound obvious, but they are the difference between finishing and not.

• Take them in order. The lessons reference each other, and Day 12 lands harder because Day 3 landed first.
• Don't binge. It is tempting, but please don't, because mental models settle overnight rather than in one sitting. Ten minutes a day is the design, not the limit.
• Read on your phone if you want. The whole thing is built phone-first, and Day 7 in line at a coffee shop is the intended experience.
• Use the glossary. It is at the top of every page, so if a word feels slippery, tap it.
• Skip the exercises if you really must. People who do them finish the course about three times more often, so make of that what you will.

Who is writing this, briefly

Girish leads product on Fidelity Investments' flagship mobile apps, built a PM operating system internally, trained hundreds of colleagues on AI Fluency, and drove adoption of AI tools at scale. Shivali is Director of Growth and Strategy at Cloudleap, where she built a set of AI workflows that automate roughly 80% of presales and sales operations, taking a prospect brief that used to require a sales rep days down to about ten minutes.

We are not professional course-makers, we are practitioners who have spent the last two years watching very smart people get stuck on things that were, in retrospect, not hard, just unseen. This course is what we wish we could have sent each of them.

Day 1 wrap

The thing you can now say plainly. Technical Enough is a 21-day course on the vocabulary and mental model AI tools assume you already have, and by the end you'll have a one-page Build Plan for the thing you want to build.

The thing you can now do. Open Day 2, where we start with the deceptively simple question: what is software, actually?

The guardrail to remember. This course teaches the map, not the craft, so don't expect to know how to write code by Day 21. Expect to know exactly which questions to ask the AI that does.

See you tomorrow.