10 · AI TRAINING TEAM

Claude Code Bootcamp

LEVEL · PRO—ENTERPRISE TEAM

The hands-on agentic coding course with Claude Code for enterprise development teams. At the terminal, tailored to your stack and your codebase. Up to 10 participants per session, on-site. Not a demo to watch, but a method to take home.

Bring the bootcamp in-house→See the programme↓

They've used our services

All our projects are covered by £10 million of professional indemnity insurance (verify here)
+ an additional £1 million dedicated to data security (verify here).

THE PACKAGES

Three formats, one method.

From a kickstart day to a full bootcamp tailored to your repo. All on-site, up to 10 participants, always hands-on. The right path depends on the team: we define it together.

Package 01

Kickstart

1 day · 7–8h

Parts 01 and 02 complete
Part 04 as a guided demo
Environment setup as pre-work
Prompt handout + templates

RECOMMENDEDPackage 02

Bootcamp

2 days · 12–14h

Complete programme, all 4 parts
Hands-on exercises for every participant
Real group exercise, all the layers
Complete material to take home

Package 03

Bootcamp + Custom

2 days + customization

The whole Bootcamp, lab on your repo
Real CLAUDE.md on your codebase
30-day follow-up included
Team .claude/ setup assistance

ADD-ONS (ON PACKAGES 01 AND 02)

Lab customization on your repo
Exercises and prompts on your real codebase
optional
30-day follow-up
½ day of review on your real PRs
optional
Internal repeat licence
Material and the right to re-run the course
on request

THE PROGRAMME · 4 PARTS

Four pillars, one hands-on training.

Four parts that build on each other: first the levers that move Claude Code's quality, then the prompt at the terminal, then teamwork, finally real code. Each part alternates a short explanation with live exercises.

Foundations

What Claude Code is and isn't. Context, prompt, verification: the three levers.

AI Dev Prompt

Mini-exercises at the terminal that produce results in the next 60 seconds.

Claude Code in a team

Roadmap, ownership, git flow, PR review, CI as a safety net.

Let's work together

Group exercise: a complete real case, all the layers, on real code.

THE THREE LEVERS

PART 01 · FOUNDATIONS

Claude Code is an agent, not an autocomplete.

Claude Code lives in your terminal.It sees the whole repo, plans before acting, runs real commands (test, git, build) and hands you back the diff — you sign off. You don't "press" it: you steer it. It's a very fast junior; you stay the senior who does code review and decides the architecture.

Quality doesn't come from magic prompts, but from three levers: context (what Claude can see), prompt (what you ask) and verification(how you check). We start here because it's the level that makes the biggest difference, on any stack.

Context = the superpower — CLAUDE.md is the most important file in the repo; attach files with @ instead of pasting; /clear when you switch task.

Atomic prompt — goal, context, constraints and a stop criterion. Verifiable in 30 seconds, not epic.

Human-in-the-loop — plan before code, diff before save, test before merge. The human sets the pace.

PART 02 · AI DEV PROMPT

Four exercises, a result in 60 seconds.

Straight to the terminal with Claude Code. Four live exercises, each designed to produce a verifiable result in about 60 seconds: a function, a test, a refactor, a fix. The point isn't speed for its own sake, but feeling the plan → code → test → verify loop under your fingers.

Eight commands give you 90% of the flow — /init, /context, /clear, /compact, /resume, /model, /agents, /review. Every participant works on their own terminal: with a maximum of 10 people, nobody just watches — everyone does.

Generate and verify — from request to tested function, in a single pass.

Targeted refactor — change structure without breaking behaviour, diff under control.

Test-driven fix — reproduce the bug, make the test pass, understand why.

LOOP AT THE TERMINAL

COMMIT → PUSH → PR → CI → MERGE

PART 03 · IN A TEAM

Claude Code in the team flow.

The same flow as always — Commit → Push → PR → CI → Merge — only Claude helps at every step and CI is the final judge. Before the human reviewer, a code-reviewer sub-agent reads the diff and flags the easy issues (missing tests, secrets, dependency rule): the human lands on a clean PR and focuses on architecture and intent.

We see how agentic coding enters the team's real work: the 5 layers of configuration, parallel git flow, PR review with AI as the first reader, and CI as a safety net — no skips, no exceptions.

5 layers — where configuration lives and how to share it across the team.

Claude does the pre-review — a dedicated sub-agent reads the diff before the human.

CI is the judge — format, test, build, arch, scan. Red CI = back to plan.

PART 04 · LET'S WORK TOGETHER

A real case, all the layers.

The most hands-on part, and the one that generates the most perceived value. In groups, on a ready demo repo, we carry a small endpoint end-to-end through all the layers — Domain, Application, Infrastructure, API and tests — with Claude Code proposing and the team verifying. The whole method from the previous parts, in a real case.

Plan mode mandatory before every change, small commits, peer review and CI as the final judge. With 10 people the hands-on is worth more than the demo: every live exercise stretches 3–4× compared to the instructor's demonstration. That's why it needs room, and why the full Bootcamp runs over two days.

Plan-first — Claude proposes file-by-file what it will create in the layers; the group approves.

All the layers — Domain, Application, Infrastructure, API and tests, on a real case.

Peer review — another group reviews, CI is the final judge.

END-TO-END · ALL THE LAYERS

OUTCOMES

What the team takes home.

Not slides, but a method the team uses from Monday morning. By the end of the bootcamp every developer can steer Claude Code on real code, with the right guardrails.

Autonomy at the terminal

Every dev steers Claude Code on their own: context, atomic prompts, plan-first.

Less rework

30 seconds of plan save 30 minutes of rollback. Diff before save, always.

Faster review

Claude does the pre-review: the human focuses on architecture and intent.

Quality protected

CI as judge: green tests, layers respected, no secrets. No exceptions.

FREQUENTLY ASKED QUESTIONS

What we get asked the most.

Transparency first. If your question isn't here, write to us: we reply within 24h, from a real person.

Who is the Claude Code Bootcamp for?

For enterprise development teams that want to adopt agentic coding with Claude Code in their daily work. It works at any seniority — from junior to lead — because the format is hands-on and adapts to the group's level. Maximum 10 participants per session, so everyone actually works at the terminal.

Is it a theory or a practical course?

Practical. Each part alternates a short explanation with live exercises at the terminal with Claude Code. In Part 04 we work together on a real case, all the layers. With 10 people the hands-on is worth more than the demo, which is why the full Bootcamp lasts 2 days.

Why Claude Code and not an autocomplete?

Because Claude Code is not an autocomplete: it's an agent that lives in your terminal. It sees the repo, plans, runs real commands, hands you back the diff — and you sign off. We teach you to steer it: context, atomic prompts, plan-first and verification. The developer stays the senior who signs and decides the architecture.

Do we work on our own code or prepared examples?

It depends on the package. The Bootcamp uses guided exercises on a sample codebase; with Bootcamp + Custom we tailor labs, prompts and CLAUDE.md to your real repo, and include a 30-day follow-up on your PRs.

What do we need to take part? Any prep beforehand?

Just a working development environment and access to Claude Code. Environment setup is handled as pre-work before the session, so the day in the room is all hands-on. We send you a checklist in advance.

How much does it cost?

Pricing depends on the package, the number of participants, the team's seniority and the level of customization. Get in touch: we'll define the right path together and send you a tailored quote.

Ready to get your team coding with AI?

A 30-minute call to understand where Claude Code can make a difference in your workflow, and build the right path. No commitment, from a real person.

Let's talk about the bootcamp→