Lapras AI Studio
How it worksCasesAboutQuestions
View as
Book a call

AI implementation tailored to your organization

Build an AI system that transforms the way your team works, and learn to run it yourselves.

Most teams unlock only a fraction of what AI can do

The real gains come from using it together. That means organizational transformation and process design: shared context, shared skills, and decisions made in one place. The system that makes it possible has three layers.

01

The personal layer

Your own system: who you are, how you work, your voice, your tools, and work that runs on its own.Each leader gets their own system: who they are, how they work, what matters now, their customers and their voice. A second brain that compounds, connected tools, skills for the work they repeat, and autonomous agents for the work that should run on its own.

02

The collaborative layer

Shared files for the team and across departments: the company, the customers, one task board.The team works from shared files: the company, the customers, the way you work, one task board. A change made once reaches everyone, across departments. This is where a leadership team stops working in parallel, and where sales, finance and operations start working from the same picture.

03

The visual layer

Dashboards per person, per department and for the leadership team: KPIs, projects and presentations in one place.Dashboards for each person, each department and the leadership team. They simplify everything to one page: KPIs, goals, priorities, open decisions and what moved this week. Project overviews and presentations are generated from the same source. The team steers from them, works on them together, and runs its meetings from them instead of from status updates.

Swipe

What's inside a personal system
Context
Who you are, how you work, what matters now. Every conversation starts already knowing it.
Instructions
How you want the work done: your rules and your standards, followed in every conversation.
Connected tools
Mail, calendar, files and your CRM, read and drafted in where you already work.
Skills and workflows
Saved ways of working you call by name, chained when one is not enough.
Autonomous agents
Work that runs on its own once the pattern is proven: triage, monitoring, follow-ups.
Security
Hard rules, your red lines, and you approve everything that leaves it.

It runs inside the tools you already use. Built in Claude Code today, made to move with you.

Apply it to every part of your business

Open an area to see what it looks like.One system, used wherever your time goes. Each area gets sharper as the others grow, because the context is shared. Open one to see what it looks like.

Sales and clients

Pipeline, account prep and follow-ups, with every relationship remembered.

  • A CRM built around how you actually sell
  • Account prep before every meeting
  • Follow-ups drafted in your voice after the call
  • A forecast the team can answer in one sitting

Leads stop slipping, because the system remembers what you would have had to.

Finance and reporting

Numbers that reconcile, ready before the meeting.

  • A reporting chain that reconciles to your accounts
  • Variance and cash views your team can read
  • Board and investor material assembled from the source
  • An obligations register that keeps itself current

The numbers are there before the question is asked.

Deal work

For advisory and finance firms: screening, materials and diligence.

  • Screening from intake to a finished spreadsheet
  • A mandate card for every live engagement
  • Client versions without hand-copied files
  • Diligence support with your own checklists

Partner time goes to judgment, not to assembling material.

Meetings and operations

Every meeting turns into decisions and tasks, without anyone typing them up.

  • Transcripts routed to the right place, with tasks extracted
  • One task board the whole team works from
  • Recurring work and follow-through caught by the system
  • Meeting prep that starts from what was decided last time

Decisions made in a meeting are still there the week after.

Content and communication

Writing that sounds like you, from first draft to the email that goes out.

  • Your tone of voice, learned from your own edits
  • Emails, decks and memos drafted in it
  • One standard for the whole team's written voice

More gets written, and it still sounds like you.

Research and strategy

Market scans and a thinking partner before the big calls.

  • Market and competitor scans on your sectors
  • Briefs structured the way you decide
  • A sparring partner before high-stakes decisions

You stay in the decision seat; the system does the legwork.

Your own day

Inbox, calendar and priorities, so you start the day knowing what matters.

  • A briefing before your day starts
  • Inbox triage: what needs you, what can wait
  • Your goals and open decisions on one page
  • Family, travel and health can live in it too, kept private

You start the day knowing what matters, not catching up on what happened.

And almost anything else you can imagine, once the system knows your company.

Built with the teams that run it today

Built with 15+ companies, from scale-ups to founder-led brands.I have built this with 15+ companies, from scale-ups to founder-led brands. Here are four of them.

Series A hardware scale-up

Organizational transformation for a team of 15

  • A personal AI system for the whole leadership team
  • A shared hub where every department works together
  • Sales and finance processes, redesigned end to end
  • One dashboard the leadership team runs on
  • An AI playbook new people onboard themselves from

M&A advisory firm

Three times the deal output, with the same team

  • A screening engine, from intake to finished Excel
  • Client confidentiality built into every step
  • Deal analysis that matched the partner's own verdict
  • 30 skills for the work the firm repeats
  • A mandate card for every live engagement

Premium glassware brand

A premium brand that went AI-native in two months

  • A B2B CRM built from scratch
  • Both Shopify stores, handled straight from Slack
  • Weekly updates and stand-ups, drafted by the system
  • Live sales signals the team steers by
  • Founder and team systems, owned by the company

Premium interior brand

Non-technical founders became builders in days

  • An accounting connector they built themselves
  • A trade-client database for their US trade clients
  • Months of store cleanup, done in three weeks
  • A brand bible and tone of voice, in their own system
  • In daily use from the first week between workshops

Swipe

How it works

Five steps to a new way of working.It starts with how your team works today, and ends with a team that works differently.

  1. Foundation

    01 Discovery

    How you work, where the time goes, what to build first.

  2.  

    02 Agreement and access

    Your red lines in writing, then access to your data.

  3.  

    03 The base

    I build each person's system from parts proven at other teams.

  4.  

    04 Workshops per person

    Hands-on, on your real work. You drive, I steer.

  5. Collaboration

    05 The team on one system

    The shared hub and the leadership dashboard, taken into use together.

  6. Deep dives, in parallel

    06 An advanced build

    Alongside the collaboration: one area in depth, such as a CRM or a forecast, with the team that owns it.

  7. Adoption

    07 Make it stick

    Every other week: new skills, new models and what is new in AI.

And it keeps going, until you run it yourselves.

How every workshop runs
  1. Inside your team. I work in your meetings, your material, your way of deciding. The system is shaped to that.
  2. Well-prepared workshops. What you need to learn next, and a system that already works when we start.
  3. Built between sessions. I build on my own initiative. You do not have to ask.
  4. Clear tasks you run yourselves. Every workshop ends with tasks ready to run on your own, so the work keeps moving between sessions, and your budget goes to the workshops.

Built, taught and kept current

01Technical architecture

A system designed around how your company works: personal, shared and visual layers, security built in, in open files you own.

02Hands-on onboarding

Workshops on your real work, and an AI playbook new people onboard themselves from. Your team learns to build, not just to use.

03A roadmap you run yourselves

Every next step written down with a ready prompt, so the work continues without me in the room.

04The latest in AI implementation

While we work together, what I learn across clients and the field is brought into your system every week or two.

Swipe

A process refined with 15+ companies, at every technical level.A process refined with 15+ companies, from first-time users to teams already deep in AI. Hundreds of hours of trial and error are already behind it, so your team starts from what works.

Security built into every layer

Three safeguards, and a person approves everything that leaves it.A system that reads your mail and your files is a security question first. Every system I build has three safeguards, and a person approves everything that leaves it.

Hard rules

Checks the AI cannot get around: no direct sending, no reading passwords.Automatic checks the AI cannot talk its way around: no direct sending, no reading passwords or keys, a scan for secrets before anything leaves the machine.

Isolation

Your files stay on your machines. Nothing is hosted by me.Your files stay on your machines. Nothing is hosted by me. Where your setup allows, the AI works in a sandbox.

Your red lines

What may never enter the system, in writing and in the system.What may never enter the system, written into the agreement and into the system itself.

Johan Stubert

From building companies to building AI systems

Johan Stubert

Founder, Lapras AI Studio

I started at BCG and spent the last decade building and scaling tech companies across logistics, e-commerce, healthcare and energy, in operations, B2B sales and go-to-market roles. Organizational transformation and systems that scale have been the thread through all of it. I run Lapras AI Studio from Stockholm.

Everything I teach, I run on myself first. My own business sits on this exact system: connected tools, daily briefings, pipeline tracking, a second brain that compounds. Every workshop, I pass on what I use every day.

I help you build your own AI system. You drive, I steer.

Common questions

What do you actually deliver?

A working AI system for each leader, a collaborative layer for the team, and the skills to keep building it. It runs inside your existing tools from day one, and everything is in files you own.

Who do you work with?

Leadership teams in companies of roughly 5 to 50 people: scale-ups, advisory and finance firms, and founder-led brands. The common thread is senior people whose time is the scarce resource.

How long does it take?

Foundation and collaboration take about two months, depending on how many people are in it. Deep dives can run alongside, one area at a time, at your pace.

Which AI do you build on?

Claude Code today. The system is built on open files rather than a platform, so it is designed to move with you to the next agentic tool.

What about data, privacy and security?

Your files stay on your machines. Nothing is uploaded to me, and model training is switched off in every account we set up. The AI drafts; it never sends or shares without your approval.

Do I need to be technical?

No. Non-technical leaders get more of the setup done for them; builders take more of the wheel. Either way, you can keep building it yourselves by the end.

What happens after?

You own the system, every file and every workflow. Nothing is locked behind my access. Most teams keep building on their own; some keep a session with me every other week.

Let's build your system

Tell me about your team and where you are on your journey.

Or tell me about your business first