AI training for employees on their real tasks.
I run AI training for employees on the tasks your team handles every week, and along with it you get written rules for what AI is and isn't allowed to do. Before the training, we go through your processes so it's clear what's worth automating and what should stay in people's hands. You can book the training on its own, with no development work, online or in person. I deliver it myself: I taught web development, UI/UX and IoT at a vocational IT school in Košice, and I build AI agents for companies and for myself.
Ing. Patrik Krajňák, Full-stack developer, Košice
How AI training on your team's tasks works
During the training, people use AI on tasks they actually have on their plate, such as replying to customers or searching through contracts. We choose the tasks together during the audit, so people try the approach on their own work and don't have to adapt it from someone else's example. We work in the AI tool your company already uses or wants to adopt, such as ChatGPT, Microsoft Copilot or Gemini.
A standard training session takes 4 hours and starts at €600. If you need to cover more people or tasks, I propose the length and number of sessions after the initial consultation. From you, I need someone who knows the team's processes, examples of typical tasks (anonymized ones are fine) and access for every participant to the AI tool we'll work in.
If you're comparing AI courses for your company, the main thing to ask is which examples your people will learn from. Here's the order I work in:
- a free consultation lasting 30 to 60 minutes, after which you get a specific price and date
- a process audit and the choice of tasks for the training
- the training itself, online or in person, as agreed
- written rules and internal procedures that you keep
How it works
How AI gets into your team's work
- 1
A real task
We pick a task your team handles every week.
- 2
The prompt
You learn to give AI a task with context and an example.
- 3
The check
You verify the numbers and sources the AI gave.
- 4
The rule
The approach goes into your written rules for what AI may and may not do.
A process audit first: what's worth automating
My work as an AI implementation consultant starts with a process audit, which shows you what's worth automating and what AI should leave alone. Together we look at where your team spends the most time, for example retyping data by hand from one system into another, and at which steps mistakes happen today.
For repetitive processes that run across several systems, an AI agent built for one specific process may be worth it. From the audit, you get an overview split into four groups:
- tasks your team can handle with an AI tool straight after the training
- processes worth handing over to an AI agent
- tasks that only a person should do, with the reasons why
- data that AI shouldn't have access to
Rules for what AI is and isn't allowed to do in your company
After the training, I give you written rules for what AI is and isn't allowed to do in your company, so the same rules apply to the whole team. Without them, everyone sets their own limits, and a client's contract can end up in a tool where it doesn't belong. If your team needs to find answers in your internal documents without anyone pasting them into other AI tools, I can build an internal knowledge assistant that works from your own documents.
I write the rules around the data you work with. They mainly set out:
- which data may go into which tool, for example personal data or contracts
- who checks the output before it goes to a client or into the accounts
- where AI only makes suggestions and a person decides
- what to do when someone isn't sure about an AI output
What your team will be able to do after the training
After the training, your people use AI on their own tasks and know when its output needs checking. You also keep internal procedures for the tasks we covered, so a new colleague can use them to get up to speed. Specifically, people will be able to:
- give AI a task with context and an example, so the output doesn't need rewriting from scratch
- check the figures and sources that AI provides
- recognize a task that AI isn't suited to
- follow the company's rules for data
EU AI Act and AI literacy: what companies must do
Under Article 4 of the EU Artificial Intelligence Act (AI Act), a company that develops AI or uses it in its work must take measures to support the AI literacy of its staff and of other people who work with AI on its behalf. The obligation has applied since 2 February 2025, and its current wording was introduced by Regulation (EU) 2026/1744, which entered into force on 27 July 2026.
The new wording of Article 4 states explicitly that a company does not have to guarantee any specific level of AI literacy for any individual. Nor does the AI Act prescribe a particular course or certificate. Training on your own tasks is one practical measure you can document: who took part and when, along with the rules the team received. If you have a specific legal question, talk to a lawyer.
Who runs the training
I run the training myself. I'm Patrik Krajňák, a full-stack developer from Košice, Slovakia, and I build AI agents and custom software for companies. From 2025 to 2026, I taught web development, UI/UX and IoT at a vocational secondary school of information technology in Košice, using teaching materials I wrote myself.
I prepare company training the same way: I write the materials and internal procedures around your team's tasks, not from a generic template. You can read more about my education and work experience on the About page.
Which work is worth handing over to AI
Building AI agents has confirmed one thing above all for me: repetitive work is only worth handing over to AI when you can describe it with clear rules.
A good agent handles on its own only what clearly fits, and when the format of the input data changes, it stops and waits for a person. A person then only checks the exceptions. The rules you get with the training work on the same principle: they set out what AI does on its own, when it has to stop and what a person must check.
Common questions
How much does AI training for teams cost?
A 4-hour AI training session for a company starts at €600. The final price depends on the scope, for example how many people and tasks the training needs to cover. You get a specific price and date after the initial consultation, which takes 30 to 60 minutes and is free.
Do we have to order AI agent development as well as the training?
No, you can book the training and process audit on their own, without any development work. If I find a process in the audit that's worth automating, I'll tell you and leave the decision to you. If you decide to go ahead, you get a quote split into parts. If you approve it or some of its parts, I build a free prototype on a sample of your data, so you can check that it will work. If you then want to continue, you pay 20% when you order and the rest after handover (see how we work together).
How can a company without an IT department implement AI?
You can start with one team and a few specific tasks, and you don't need an IT department to do it. As an external AI consultant, I prepare everything from the process audit to the rules for working with data, and I train people directly on their own tasks. If you later want to connect AI to your systems, I'll build the technical part as well and document it.
Do you run AI training in person, or only online?
I train both online and in person, depending on what suits your team better. I'm based in Košice, Slovakia, so in-person training in and around Košice is the easiest to arrange. For companies elsewhere in Slovakia, in Czechia or further afield, we agree on the format together.
Does the EU AI Act require AI literacy training for employees?
The AI Act doesn't prescribe any specific training or certificate for employees. A company that uses AI has to take measures to support the AI literacy of the people who work with AI, and training is one such measure. It can be documented and ties in with the rules your team follows when working with AI.
More services and projects
- Service AI agents for business Custom AI agents process emails, orders and documents using fixed rules, and a person approves any step involving money. The consultation is free.
- Service AI knowledge assistant An internal AI chatbot that answers staff questions from your contracts, procedures and meeting notes, with sources and access rights. Free prototype.
- Service Business process automation Business process automation for small and mid-sized companies: invoices, documents, quotes, orders, stock, attendance and reports. One process from €500.
- Service Custom software development Custom software from Košice: ERP, internal systems, web and mobile apps, integrations. Experience at U.S. Steel and GymBeam. Free prototype before you order.
Contact
Tell me what you still do by hand.
A few lines about the process that slows you down or the system you need are enough. Within eight hours I'll get back to you on whether and how it can be built, and if it makes sense, we'll set up a short call.
Prefer email or phone?
patrik.krajnak.dev@gmail.com +421 950 625 466
I reply within 8 hours. Based in Košice, Slovakia (CET), working remotely
Calls on Google Meet: Monday to Thursday 7:00 to 20:00, Friday until 18:00. Weekends by arrangement.
I use your details only to reply and to arrange the call. Privacy policy