Your team is using Claude for autocomplete.
Your competitors are shipping with it.
One day. Your office. Your codebase. I walk in at 09:00 and your team walks out at 16:00 having shipped three real things with Claude Code.
Most AI pilots
never leave the pilot.
“Fewer than 5% of enterprise AI pilots deliver measurable business value.”
I spend most days shipping production code with Claude. Not demos. Not slides. Commits that go to main. The gap between “my team tried ChatGPT” and “my team ships with AI” isn't a tools gap. It's a craft gap.
You can't close a craft gap with a webinar. You close it by sitting next to someone who does the thing you're trying to learn, with your real code on the screen, for one concentrated day. That's what this is.
Real output.
Not a certificate.
No one needs another AI course completion badge. Your team needs working systems, running in your repo, that they built themselves.
- Ship three working things — a real agent, a code-review workflow, and an automation of your team's choice — all merged to your main branch that day.
- Leave with the Claude Ship Playbook — your team's own prompt library, MCP configs, and repo conventions, captured live during the workshop.
- Master parallel subagent workflows — dispatch five Claudes on five tasks, review, merge; do a week of work by Friday.
- Wire Claude Code into your review loop — PR generation, review, and triage patterns your reviewers actually trust.
- Build an eval harness for your agents — so you can defend every decision, not just the wins you remember.
- Walk away with a 30-day rollout plan — a working document for extending Day-1 wins to the rest of engineering.
The Claude
Ship Playbook.
The Playbook is your team's specific one. Not a generic PDF.
During the workshop we build a living document containing your team's prompt library, your Claude Code skills and slash commands, your MCP configurations, your agent design patterns, and the exact hands-off workflows we validate live.
You leave owning it. I don't.
Part of every engagement · Ships yours by end of day
- 01Prompt library — voice, patterns, guardrails
- 02Claude Code skills & slash commands
- 03MCP configurations for your stack
- 04Agent design patterns
- 05PR review workflows
- 06Eval harness conventions
- 0730-day rollout plan
09:00 to 16:00.
Your office. Your code.
Every hour accounted for. No filler. No icebreakers. No slide theatre.
Opening & codebase walk-through
Your tech lead, your stack, your three wedge wins for the day.
Claude Code foundations, live
Your first agent, your repo conventions, your first merged PR.
Break
Coffee. Breathe. Talk to the person next to you.
Parallel subagents
Dispatching five Claudes across five tasks on your real backlog — review, merge.
Working lunch
Agent design canvas, paired. Food provided.
PR review workflows
Claude as reviewer, reviewer-of-reviewers, triage patterns your reviewers actually trust.
Eval harness build
So your team trusts what ships, not just the wins you remember.
Playbook capture + 30-day rollout plan
Everything we built today, written down. Your team owns it.
Not every team.
The right teams.
Right fit
- Engineering teams shipping real production code.
- Technical leads comfortable giving Claude repo access on-site.
- 6–12 engineers in the room.
- Companies ready to change day-to-day habits, not just buy licenses.
Not a fit
- Purely strategic AI-literacy workshops for executives.
- Beginner ChatGPT or prompt-engineering courses.
- Classroom settings without access to real codebases.
- Teams looking for a certificate to put on LinkedIn.
Jonas Hulthén.
Ex-Nordnet software engineer. Co-founder of Webnation, where I ship production code with Claude every single day. Public speaker on growth mindset, psychological safety, and collaboration — because rolling out AI well is a team-behaviour problem before it's a tools problem.
This workshop exists because I got tired of watching companies run pilots that never shipped. I know what the gap looks like from the inside.
The day works because I do this myself every day. I'm not teaching you what I read. I'm teaching you what I ship.
Flat fee. Your team,
not per-seat.
Per-seat pricing signals "training vendor." Flat fee signals "partner." All prices ex VAT, ex travel.
Pilot / Founding cohort
ex VAT, ex travel
List price 125,000 SEK
Founding-client pricing for the first three clients in exchange for a case study.
- 1 day on-site
- 6–10 engineers
- Claude Ship Playbook + 30-day rollout plan
- One 90-min clinic 2–4 weeks post-workshop
- Pre-workshop intake call with your tech lead
Ship-with-Claude Program
ex VAT, ex travel
The full programme. One day in the room, two clinics to keep the rollout alive.
- 1 day on-site + 2 half-day virtual clinics (week 2, week 6)
- 8–12 engineers
- Claude Ship Playbook + 30-day rollout plan
- 60 days of private Slack / Teams access
- Loom walkthrough of three wins shipped during the day
Org Transformation
ex VAT, ex travel
Multi-team rollout across 12 weeks. Champion programme included.
- Day-1 bootcamp + 2 additional team days
- 20–40 engineers across multiple teams
- 5-person AI champions cohort
- Weekly 2-hour leaders' clinic (8 weeks)
- Procurement/security write-up for your CISO
- 3-month retrospective workshop
Nordic corporates typically expense this from engineering or developer-experience budgets. Happy to provide an approval letter for your manager.
Frequently asked,
honestly answered.
If your question isn't here, email me directly at jonas.speaking2@gmail.com. I read every one.
Why one day? Isn't this a multi-week programme?
I read your blog and watched your Looms. Is the workshop different?
Can we run this for 20 people?
Can we expense this from L&D?
Do you record the day?
Languages?
What if we cancel?
Why Claude specifically? Why not Cursor or Copilot?
Run the bootcamp
at your company.
Tell me about your team, your codebase, and what you want shipped by 16:00. I reply within one business day.
Book a discovery calljonas.speaking2@gmail.com · Stockholm, Sweden