AI training for teams in companies of 10 to 200 people
Your staff already use AI. Let’s teach them to do it well and safely.
Hands-on workshops for accounting, sales, customer service, administration and the people who put reports together. On your documents and in your tools, without company data ending up where it should not. Run by a developer who works with AI every day.
- Up to 12 people per group
- On your own examples
- 30 days of support afterwards
- Reply within one business day
Does this sound familiar?
- Staff use ChatGPT on their own and you do not know what data goes in.
- You bought Microsoft 365 Copilot, Gemini or ChatGPT for business and nobody uses it.
- You hear that AI speeds up accounting and sales but do not know where to start.
- Every month someone assembles a report by hand from several files and systems.
- Generic “AI courses” have nothing to do with how your department works.
- Someone mentioned the AI Act and an obligation to train staff, and you are not sure what it means for you.
Modules
Each module is a workshop built on examples from that department’s work. You pick the modules. The opening module is shared.
0. Safe AI at work
Opening module for everyone, 1–1.5 hours. What AI does well and where it makes things up. What data you can paste and what you cannot. Personal versus business accounts. Rules for the company and the AI Act obligation.
1. AI in accounting
Reading invoices and bank statements, reconciling two lists, e-invoicing and tax reporting files in a spreadsheet, formulas written with AI, draft letters. The limit: AI does not make tax decisions.
2. AI in sales
Call preparation, emails and proposals in your company’s voice, meeting notes straight into the CRM. AI in Salesforce and other CRMs, and the data hygiene without which it does not work.
3. AI in other departments
Customer service, administration and HR, marketing, management. Your choice, depending on who is in the room.
4. Reporting and data
Plain-language questions to your data, SQL and formulas written by AI and verified by the participant, Excel, Power BI, Looker Studio, a read-only MCP connection, a recurring report.
After training: implementation
The workshop ends with a list of ideas rated “you can do it yourselves” or “needs a developer”. The latter I can build. A separate site about AI implementation.
Why a developer, not a trainer
I show how to connect AI to data
Not just how to write prompts. A query to a database, a report from the CRM, a connection through MCP: I do this at work and I will show it at yours.
I know where AI gets it wrong
I read AI-generated code and output every day. I teach people to check, not to trust. That is the biggest difference between an employee who gains from AI and one who loses.
I was there for a company rollout
Finance, accounting, administration, IT at my employer. I know what an accountant asks and what the head of IT worries about, because I heard it in person.
I take permissions seriously
AI sees whatever the account it runs on sees. Before anyone connects an assistant to the CRM, I show how to do it on a separate, read-only account.
On the tools you already have
Claude, ChatGPT, Copilot or Gemini. I do not sell licences or a platform of my own. If you already pay for something, I teach on that.
I stay after the session
30 days of questions by email and a two-hour session after a month. And if the workshop produces something worth building, I can build it.
In numbers
- 6 yearson the Salesforce platform, since 2020
- 4Salesforce certifications
- 3 countriesPoland, Germany, Sweden: companies I have worked for
- 5,000+active users of my Android apps
- up to 12people per group, so everyone gets to work on their own tasks
- 30 daysof email support after the training
How a training is built
A call about your needs
What the department does, which tools it uses, what eats the most time. 30 minutes, online.
A survey of participants
Short, before the session: role, tools, 2–3 tasks they would like to do faster.
A programme tailored to the department
I pick the modules and exercises. You get the programme in writing to approve, broken down by the hour.
A workshop on your company’s examples
Everyone works on their own tasks, in groups by role. I move between desks instead of standing by the slides.
Materials and 30 days of questions
Ready-made prompts for the department, AI usage rules, a list of ideas to implement. Then email support and a session after a month.
Where my examples come from
I do not show slides from other people’s projects. The exercises in a workshop come from three places.
My own daily work
I work with AI every day: Claude Code as a developer, Claude and OpenAI in automations that run in production, Agentforce and Prompt Builder in Salesforce. I built my own MCP server with around 75 tools and use it daily, so I know what happens when an AI assistant gets access to real data.
The AI rollout at my employer
At the company where I own Salesforce, I took part in rolling out AI in finance, accounting, administration and IT. I saw which tasks genuinely got faster, where people stopped using the tool after a week, and what questions accounting asks before pasting anything into a chat. The team I work in switched from ChatGPT to Claude.
Your participant survey
Before a workshop every participant describes 2–3 tasks that take up most of their time. Those become the exercises. That is why a workshop for accounting looks nothing like one for sales, and yours looks nothing like the company next door’s.
What you will not find here
Numbers of people trained or participant testimonials. Training for companies is a new service for me, built on a developer’s experience and on the rollout at my employer. I say so plainly rather than inventing references. After the first workshops, descriptions will appear here with the clients’ consent.
A developer or a training company?
For a team that is supposed to start using AI at work, what matters is whether anyone actually works differently afterwards. The rest is detail.
| A developer running the workshop (me) | Training company | Online course | Nothing | |
|---|---|---|---|---|
| What you practise on | Your own tasks from the survey, in your own tools | The trainer’s examples, often the same for every client | The examples in the recording | — |
| Company data and safety | Usage rules ready for management to adopt, anonymisation done live | Usually one slide on GDPR | Usually skipped | Everyone does their own thing on a personal account |
| Reporting and connecting to data | SQL, formulas, MCP, a recurring report: I show it because I do it | Rarely; the trainer is usually not a developer | No | — |
| After the session | 30 days of questions by email, a session after a month, a list of things to build | A feedback form and a certificate | Access to the recording | The tool sits idle, the licence keeps billing |
| Building what came out of the workshop | I can do it myself: a report, an integration, AI in the CRM | You need to find a contractor | You need to find a contractor | — |
| Group size | Up to 12, everyone works on their own tasks | Varies; 30 in a room is common | Unlimited, no contact | — |
From the knowledge base
Short pieces on what I hear most often from companies before a training.
AI Act, Article 4, in a small company
If anyone in your company uses ChatGPT, Copilot or Gemini at work, this provision applies to you. It is not as frightening as it sounds, but it does not go away if ignored.
Five accounting tasks with AI
Accounting is a lot of repetitive work with documents and reconciliations. AI can shorten it, provided the accountant checks the result and the decisions stay with her.
Copilot bought, nobody uses it
Microsoft 365 Copilot, Gemini in Google Workspace, ChatGPT for business. Licences paid, everyone clicked something in the first week, a month later two people use it. This is not a problem with the tool.
Who runs it
I have been building on Salesforce since 2020. Today I am the sole owner of the Salesforce org at a Swedish software company: development, permissions, integrations, production releases. Before that, two years in a Vaillant Group team (loyalty programme, large Apex API, installer portal) and three years at Britenet for clients in education, insurance and finance. Salesforce certifications: Platform Developer I, Administrator, Platform App Builder, Sales Cloud Consultant.
I work with AI every day: Claude Code as a developer, Claude and OpenAI in production automations, Agentforce and Prompt Builder in Salesforce. I built my own MCP server with around 75 tools, exposed to the Claude assistant over OAuth 2.1, and a system where AI wrote nearly all of the code while I owned the specification, review and tests. At the company where I own Salesforce, I took part in rolling out AI in finance, accounting, administration and IT.
After the workshop you have a list of things to build. If you want, I build them.
A recurring report, AI connected to your CRM or database, email and documents. A separate site about AI implementation.
Ask about training for your team
Tell me the department and the headcount. I reply within one business day with questions or straight away with a proposed programme. Quoted after a call, depending on format and headcount.