Claude for business,
in writing.
Notes from running a one-person practice.
By role
Four AI agents or one? I built the same quote desk twice
Getting started with Claude in your team: a plan for the first 30 days
Three AI sales agents, one shared memory: 215 corrections fell to 92
Claude Opus 5.5 fast mode: the answer on the call came in 2.7 seconds, not 6.6
Claude's morning brief said "no new deals". Two had been won on Friday afternoon
Claude in Google Sheets can clean up your CRM. Read the merges first
Claude Haiku 5.5 got 9 of 300 enquiries wrong. You can find them all by reading 10
Claude Code mods: a CPR filter that tripped over a Danish letter
Claude wrote the renewals brief. One line from the team lead turned 7 into 12
Claude Sonnet 5.5 can make your agent go quiet. Here's a checker
Claude Sonnet 5.5 vs Opus 5.5: I tested both on the same sales forecast
Claude Tag can now use your own connectors. The whole channel sees the answer
Everyone can build an agent now. Nobody has signed off on them
Claude Opus 5.5 shipped with nine benchmarks. Anthropic hedged them itself
The block became a message. Who gets it where you work?
Claude Cowork and chat become one. Your internal guide is now wrong
43 workflows from Claude. How many of them reach a Danish company?
Your AI key may act on nobody's behalf
A better model, a 36% bigger bill, and an unread prompt
AI workflow automation: what to build first
AI hallucinations: the model guesses because silence scores zero
Claude can block a prompt. Your timeout decides the rest
What does an AI consultant cost in Denmark?
Eight features fall outside your zero data retention arrangement
Claude leaves two kinds of mark. One of them you can check today
Word, Excel, PowerPoint, Outlook. One tick now covers four more programs.
Claude commerce agents: the three things you can check yourself
Anthropic MHS: integration took weeks. Now it takes hours. Then the hard part started.
Claude Fable 5.1: seven differences nobody deployed
The line for what Claude may do on its own has to be written in advance
Claude memory has shipped. The default depends on what you pay for.
Claude now marks its own output. It proves less than you think.
Anthropic now scans skills. Five piles, one gets checked
Eight seats at the same table: Anthropic's sharpest hire
What six course certificates in five weeks actually changed
EU AI Act: what your company must do now the transparency rules apply
Safe AI implementation: three incidents at Anthropic, and what they say about your next agent
Claude Opus 5: same price, a much bigger leap in capability
Claude can now show you how you use AI. And why it is about being deliberate, not using more
Claude can now work inside your Microsoft 365, not just search it
Claude Cowork on web and mobile: workflow automation that runs without your computer
Alberta secured 466 million lines of code with Claude in 20 hours
Anthropic just published what Fable 5 blocks on cybersecurity. Here is how to turn it into a policy that survives an audit
Claude Sonnet 5: top-model quality for business, at a far lower price
Fable 5 is back worldwide: the lesson is resilience against model dependency
Claude Tag in Slack: Claude integration as a permanent team member for Nordic companies
Fable 5 and Mythos 5 were shut down without warning. What to do now, and why it is about dependency
DXC is building Claude into banks, airlines and insurance. What the order means for your company
Anthropic wants to let governments stop dangerous AI models. What it means for you as a leader
A resume that gets you the job.
What Claude actually does inside a 12-person SaaS team.
Why I only work on Claude.
Get new essays by email.
Roughly twice a month. Same voice. No list rental, no retargeting.
Claude for business,
put to work.
Reading is the easy part. Tell me what you're trying to ship and I'll tell you whether Claude is the right tool, honestly. Either way, you leave with a plan.
You pick a slot in Kim's calendar. No form, no callback.
Certified by Anthropic · Fixed price on every stage · Invoiced in DKK