Welcome & who this is for

You can ship features all day. Then someone says "design X for millions of users" and your mind goes quiet. This course closes exactly that gap, and it's the gap between Senior and Staff.

You're not missing knowledge. You're missing a method

• What you can already do Build the feature, wire the endpoint, ship it, debug it in prod. Real, hard-won craft.

so why the freeze?

• Where it stalls "Design a URL shortener for 100M users." Blank. Not because you don't know things, because you have no way to reason from them.

The blank isn't a knowledge gap. It's the absence of a repeatable way to think under an open-ended prompt. That's a learnable skill, and it's the whole course.

If any of these is you, you're in the right room

• The Staff-track senior You do the work of a senior and you're being told to "operate at the next level." Nobody has said what that concretely means. This does.
• The interview candidate You have a system-design or architecture round coming and you don't want to memorise 20 diagrams and pray one shows up.
• The builder who freezes You can build anything concrete but "design a system" feels like a different, scarier job. It isn't. It's the same reasoning, one level up.

One thing ties all three together: you can build, and now you need to decide and defend.

Two people should close this tab now

✗ The framework-tutorial hunter

If you want "React vs Next.js" or "how to set up Kafka," this isn't it. Tools change every 18 months; the reasoning here doesn't. We name tools only as examples of a trade-off.

✗ The copy-paste architect

If you want twenty reference diagrams to memorise and reproduce, you'll be brittle the moment the prompt shifts. We're doing the opposite: derive the diagram from first principles, live.

Memorised architectures survive until the interviewer changes one constraint. Reasoning survives anything.

By the end, you can design any large system, and defend every trade-off out loud

• Producing a design is table stakes. Defending it out loud, "I chose this, here's what it costs, here's when I'd change it," is the actual signal. It's what a promo committee looks for and what an interviewer scores.

Anyone can draw boxes. The staff engineer says why these boxes, what they trade, and what would make them wrong. That sentence, on demand, is what you'll leave with.

Reason from a few axes, don't memorise twenty architectures

• A small set of axes Consistency ↔ availability, latency ↔ throughput, and a handful more. Every system is a position on these.
• One idea per deck Each lesson is a single animated deck on exactly one idea. No 90-minute firehose. Watch, step, done.
• Derive, don't recall Given any prompt, you place it on the axes and the architecture falls out. New prompt, same method.
• Say it out loud Every axis comes with a way to name your position, so the defense is built in, not bolted on.

This is the method in one line: you don't start from a remembered diagram, you start from the prompt, place it on the axes, and let the architecture and its defense follow.

Twenty memorised architectures cover twenty prompts. A handful of axes covers every prompt, including the ones nobody's written a blog post about yet.

Not just watched. Built

• 6 graded artifacts Real designs and ADRs you produce and get feedback on. This is where the learning actually sticks, not in the watching.
• 12 cheat sheets One-page references for each axis and pattern. The thing you skim the night before the interview or the design review.
• 8 worked designs Full systems reasoned from the axes end to end, so you see the method carry all the way through, not just in slogans.

You finish with a body of work you can point to, not a certificate you can't.

The interview and the promo reward the same skill

• The interview loop The system-design round isn't testing recall. It's watching whether you can reason under ambiguity and defend a call. Same skill.
• The promo case "Operates at Staff level" means: makes and owns trade-offs others trust. You can't fake it; you can practise it. Here.

You're not preparing for two different things. You're building one muscle that both the loop and the ladder are looking for.

Build you already have. This course adds decide, and defend

1. It's for you if you're a senior aiming at Staff, or facing a system-design round, and you freeze at "design X for millions."
2. The promise: design any large system and defend every trade-off out loud, the part that actually gets scored.
3. The method: reason from a few trade-off axes, not twenty memorised architectures. One single-idea deck at a time.
4. Next: 0.2 · How to use this course + the workbook, so you get the most out of the decks, cheat sheets, and graded artifacts.