AI Training for Businesses in Latvia

What stage is this offering at?

To be upfront: the program and the price have not been finalized yet. We are developing it, and we do not want to sell something that is not ready. This page is published because the question of AI training for businesses is topical right now, and it is worth knowing where things stand.

If your team needs training now, write to us — we will tell you what stage the program is at and agree on a format suited to your situation. We will not promise a number of hours, a price or a certificate that we cannot yet confirm.

Why is AI training a topic right now?

Since 2 February 2025, Article 4 of the European AI Regulation (AI Act) has imposed an obligation regarding AI literacy. It applies not only to developers, but to any organization that uses AI systems — even when "using" means an accountant and a sales manager opening ChatGPT.

In July 2026, the Digital Omnibus Regulation (EU) 2026/1744 entered into force, which softened the wording: the development of AI literacy has to be supported, rather than a specific level guaranteed. The article was not repealed, and the date was not moved — from 2 August 2026, national supervisory authorities gain formal powers to oversee this area.

The practical meaning is simple: there is no certificate to buy and no state program, but you need to be able to show that your AI users are trained appropriately for their role, and that there is a record of it. We have written up what that means step by step separately: AI Act Article 4 and AI literacy. (This is not legal advice.)

Who will lead the training?

The training will be led by Levs Fainglozs, Dr.oec. — co-founder of learnxMI. learnxMI is a joint project of Garri Gelfers and Dr.oec. Levs Fainglozs.

• A lecturer since 2007 — at the Transport and Telecommunication Institute and RISEBA University of Applied Sciences. Almost twenty years of work in the classroom, not a course put together last year.
• 16 academic publications, 55 citations. Research areas: business manageability and processes, turning goals into results, and the criteria investors use to evaluate a company.
• In 2024, co-author of the study "Employers' Expectations of Students' Generative AI Skills" — on which generative AI skills employers actually expect. The program will be built directly on this question.
• A RISEBA Certificate of Honor (2017) for his contribution to research on deferred corporate income tax as a growth factor for the Latvian economy.
• The profile can be checked independently: ResearchGate.

How do you choose an AI trainer?

The AI training market in Latvia is new: it includes both experienced practitioners and people who put together their first course a few months ago. This is not a list of accusations — these are signs you can check yourself, even if you end up choosing someone other than us.

• Is there a verifiable profile? A name, an affiliation, publications you can open on ResearchGate, Google Scholar or ORCID. If the only evidence is a website, that is a description, not proof.
• Has the person taught before the AI wave too? Teaching is a separate skill. Knowing a tool and being able to teach it to twenty employees of different ages are not the same thing.
• Is the program adapted to your industry? Ask which of your own documents the sessions will work with. If the program is the same for everyone, the team will be learning from someone else's examples.
• Does it cover security and data, not just prompts? The question of "what is OK to put into the chat" costs a company the most when it goes unanswered.
• Does the trainer stay reachable after the course? Questions do not come up during the session, but in the third week, in the middle of a real task.

What will the program be built around?

We will confirm the exact content and format, but the direction is set, and it is no secret.

• On your real tasks — price lists, letters, reports, documents. Not on demo examples from a slide deck.
• By role — a manager who makes decisions and an employee who uses the tool every day need a different depth of training.
• With a security component — what may and may not be entered, where data protection risk arises, how to check an AI answer.
• With documentation — the program, the list of participants and an attendance record: exactly the records that let you show the training took place. (This is not legal advice.)
• In Latvian, Russian or English, so the whole team can learn, not only the part that speaks one language.

Why are training and implementation in one place?

The usual scenario: the course happens, people get inspired, the trainer leaves, and a month later everything is back to the old routine, because nobody built what was discussed in the course.

For us, training is one half; the other is building systems: workflows and AI agents that connect email, CRM and internal tools. An idea found during the course, "we do this by hand every week," can become a built process instead of a note in a notebook. The free AI opportunity audit shows where to start.