A good bootcamp isn't a longer workshop — it's a designed progression from 'I've heard of this' to 'I can do this'. Here's the process behind designing one that genuinely builds capability, the same approach used for kenai's 3-day AI bootcamp.
Start from outcomes, not topics
Don't list what you'll cover; list what learners will be able to do at the end. 'Architect a solution with Claude', 'build with Claude Code', 'ship an automation'. Every session then exists to serve a concrete capability.
Sequence from foundation to build
Capability stacks. Day one builds the mental model (how the tools work and how to architect with them). Day two moves to building hands-on. Day three composes those skills into real automation. Each day depends on the last.
Make every session hands-on
If a session has no build, cut it or redesign it. Learners should leave each day having made something — that's what converts knowledge into skill and confidence.
End with a capstone
A capstone forces integration: the learner applies everything to a real problem of their own. It's both the best assessment and the strongest proof of capability — something they can show, not just a certificate they hold.
Design for life after the bootcamp
Send people home with reusable assets — prompt libraries, templates, automation blueprints — and a plan to keep building. The goal is durable capability, not a good three days.
The test of a bootcamp isn't how it felt on day three. It's what people are still doing differently a month later.
This is exactly how the kenai 3-day bootcamp is structured: Day 1 Claude architecture, Day 2 Claude Cowork & Claude Code, Day 3 automation with Codex and agents — finishing with a capstone.