encorp.ai runs AI training for companies and teams in Bulgaria — practical programmes for Microsoft Copilot, ChatGPT, Claude and Gemini, by department and role. The training is the core of phase 1 (Enablement) of the Encorp AIM method — an entry to the process, not an end goal — and ends with a certificate of completion of AI literacy training (Article 4 of the EU AI Act).
Hands-on training for teams — the core of phase 1 (Enablement) of Encorp AIM
6 modules · 20 hours · certificate at the end
Trusted by leading companies
See the portfolio →











| Who it is for | Teams of up to 20 people; large companies by department |
|---|---|
| Format | On-site at your company or online |
| Duration | 4 hours (one session) or 20 hours (a 5-session programme) |
| Language | Bulgarian or English |
| Price | from €1,200 excl. VAT |
| Encorp AIM phase | Phase 1 — Enablement (weeks 1–4): training by department |
| Deliverable | A certificate of completion of AI literacy training (Article 4 of the EU AI Act) for every participant |
AI training for companies covers the fundamentals of working with AI, hands-on tasks from the team's daily work, and the rules for handling data safely. We start with what language models can and cannot do, when to trust them and when to check the result. Participants then work with the tool the company uses — Microsoft Copilot, ChatGPT, Claude or Gemini — on tasks from their own department.
Every group also goes through a security module: what never goes into an AI tool, how to recognise an attempt to manipulate the model, and what the company's internal AI policy says. Exercises use your anonymised data so they stay close to real work. At the end every participant receives a certificate of completion of AI literacy training (Article 4 of the EU AI Act); the 20-hour programme also includes an online exam.
The training runs in three steps: a needs call, delivery, and a check of what was learned. In the first call we agree which departments take part, which tools and licences you already have, and which tasks you want to speed up. We build the examples and exercises around that — your processes, not generic demos.
The starter session is one 4-hour session. The team programme is five 4-hour sessions scheduled around the team's calendar — for example once a week, so people can apply what they learned between sessions. Groups are capped at 20 people to leave time for practice and questions. We deliver on-site at your company or online, in Bulgarian or English. After the programme you receive a summary of what was covered and the next steps for the team.
AI training is the core of phase 1 of the Encorp AIM method — Enablement (weeks 1–4). In it, teams are trained by department, we prepare a readiness profile and an opportunity map, draft an AI policy, and by week 4 the team has a working quick win. Phase 2, Roadmap and PoC (weeks 5–10), and phase 3, Deployment (weeks 11–12), follow, with two decision gates in between. The three phases and the packages are described on AI transformation for companies · the Encorp AIM method.
Training is an entry to the process, not an end goal: a team that knows how to work with AI is the starting point, and the result comes from processes that change. After a stand-alone training the usual next step is a meeting with an AI Director, where we pick the first quick win and the scope of the 12-week AI transformation. One transformation covers a team of up to 20 people — the same size as a training group — and the Fractional AI Director owns the result end to end. Strategy and the roadmap are part of the first team's transformation, not a separate paid step. The training can also be booked on its own.
The training is built around the requirement in Article 4 of the EU AI Act that companies ensure a sufficient level of AI literacy among staff who work with AI systems. It explains how the tools work, what the risks are and how to use them responsibly in your company. Every participant receives a certificate of completion of AI literacy training (Article 4 of the EU AI Act), which the company keeps in its records.
The certificate documents the training; on its own it is not a legal assessment of compliance. What exactly a company has to do under Article 4 is explained in the guide AI literacy under Article 4 of the EU AI Act. For policies, the risk register and oversight of AI systems, see AI governance and the EU AI Act. This information is not legal advice.
We train on the tool the company already uses or has chosen, because the fastest effect comes from working with the real licences. If the company works in Microsoft 365, Microsoft Copilot is usually the logical choice; in Google Workspace, Gemini. ChatGPT and Claude are often chosen by teams that do a lot of writing, analysis and programming.
If the choice is still open, the comparison in Microsoft Copilot, ChatGPT Enterprise or Claude: which fits which company helps you decide. Encorp AIM also has two tracks: the Microsoft Copilot track for all employees who work in Microsoft 365, rolled out department by department, and a tool-agnostic track with ChatGPT and Claude for managers and advanced users. Copilot is not required, and many companies combine the two — Microsoft Copilot for everyone and ChatGPT or Claude for a smaller group. Details for each tool are on the training pages below.
| Format | Duration | Participants | Language | Location | Price (excl. VAT) |
|---|---|---|---|---|---|
| Starter session | 4 hours | up to 20 people | Bulgarian or English | on-site or online | €1,200 |
| Team programme | 20 hours (5 sessions of 4 hours) | up to 20 people | Bulgarian or English | on-site or online | €5,000 |
| Company-wide programme | agreed per company | several teams | Bulgarian or English | on-site or online | on request |
Prices exclude VAT. For a company-wide programme with several teams we prepare a quote after a call.
The fundamentals are shared; the hands-on tasks come from the work of each department.
Meeting preparation, emails and proposals to clients, call summaries and CRM notes.
Job ads, candidate screening, internal rules and employee communication — with care for personal data.
Spreadsheet and report analysis, reconciliations, variance explanations and reports for management.
Writing and reviewing code, documentation, incident handling and the safe use of tools such as GitHub Copilot.
AI strategy and priorities, tool choice, risks and EU AI Act requirements — and what to ask your teams.
Unternehmen bezahlen für Copilot, ChatGPT Enterprise oder maßgeschneiderte KI-Agenten — und ein Quartal später bleibt ein großer Teil der Lizenzen ungenutzt. Entweder wirken die Tools einschüchternd, oder die Mitarbeiter vertrauen ihnen bei echter Arbeit nicht, oder Schatten-KI übernimmt, weil es keine Richtlinie gibt. Training ist die Lücke. Richtig gemacht, verwandelt es KI von „diesem Tool, für das wir bezahlen“ in einen Multiplikator für jede Rolle.
Unsere Trainings sind rollenspezifisch, praxisorientiert und mit echten Unternehmensszenarien. Wir schulen Ihr Finanzteam zu finanziellen KI-Aufgaben, Ihr Vertriebsteam zur KI-Nutzung im CRM, Ihr Engineering-Team zu Code-KI — nicht dieselbe generische Präsentation für alle.
Separate Tracks für Führungskräfte, Operations, Vertrieb, Marketing, Engineering, Finanzen, HR. Gleiche Grundlagen, unterschiedliche praktische Beispiele.
Echte Unternehmensdaten (anonymisiert), echte Aufgaben aus dem Arbeitsalltag. Am Ende des Tages haben die Teilnehmer selbst etwas auf ihrem eigenen Laptop automatisiert.
Jeder Track vermittelt Datenklassifizierung, Anbieterrichtlinien, Bewusstsein für Prompt Injection und Ihre spezifische KI-Nutzungsrichtlinie.
Praktische Bewertung je Track und eine Teilnahmebescheinigung über eine Schulung zur KI-Kompetenz (Artikel 4 des EU AI Act) für jeden Mitarbeitenden — nützlich für Leistungsbeurteilungen und die Unterlagen des Unternehmens.
Trainer unterrichten auf Bulgarisch und Englisch. Materialien in beiden Sprachen. Internationale Muttergesellschaften können dasselbe Training auf ihre bulgarische Tochtergesellschaft übertragen.
KI entwickelt sich monatlich weiter. Wir bieten vierteljährliche Auffrischungskurse, asynchronen Q&A-Zugang und eine Praxisgemeinschaft unter Ihren geschulten Mitarbeitern.
Sechs Module, insgesamt ca. 4½ Tage pro Track, verteilt auf 2–3 Sitzungen über 2–4 Wochen (nicht zusammengepresst in eine einzige, ermüdende Woche). Jedes Modul endet mit einer praktischen Übung und einem Checkpoint — am Ende hat jeder Teilnehmer die Arbeit selbst gemacht, nicht nur davon gehört.
Was LLMs sind und was nicht. Fähigkeiten, Grenzen, Halluzinationen, wann KI hilft und wann nicht. Kein Hype, keine Angstmacherei.
Praktische Prompt-Muster für den Arbeitsalltag. Vorlagen, die das Team tatsächlich wiederverwenden wird. Die relevanten Unterschiede zwischen ChatGPT, Claude und Gemini.
Vertiefung je Track zu den für die Rolle relevanten KI-Tools — Copilot für Engineering, GPT-Sales für Vertrieb usw. Praxisnah mit anonymisierten Unternehmensdaten.
Datenklassifizierung, Anbieterrichtlinien, was NICHT in ChatGPT eingefügt werden darf, Bewusstsein für Prompt Injection, die KI-Nutzungsrichtlinie Ihres Unternehmens.
Abschlussprojekt: Jeder Teilnehmer baut einen kleinen KI-Agenten, der eine echte Aufgabe aus seinem Job automatisiert. Sie gehen mit etwas Selbstgebautem, das sie nutzen, nach Hause.
Praktische und schriftliche Bewertung. Jeder Teilnehmer erhält eine Teilnahmebescheinigung über eine Schulung zur KI-Kompetenz (Artikel 4 des EU AI Act) für die HR-Unterlagen und für Audits.
Wir passen jede Kohorte an das Unternehmen an. Die Bedarfsaufnahme vor dem Training legt fest, welche Tools und Workflows im Fokus stehen; die Community nach dem Training hält die Mitarbeiter auf dem Laufenden.
Die meisten Engagements laufen für eine einzelne Kohorte insgesamt 4 Wochen. Rollouts mit mehreren Kohorten werden je nach Unternehmensgröße geplant.
Interviews mit Stakeholdern. Tool-Bestandsaufnahme. Rollenanalyse. Wir passen die Modulinhalte an Ihren Stack, Ihre Daten und Ihre Richtlinien an.
Grundlagen, Prompting, rollenspezifische Tools. Mischung aus Präsenzworkshops und asynchronen Übungen.
Sicherheit & Richtlinien, danach das Abschlussprojekt: Jeder Teilnehmer baut einen kleinen KI-Agenten, der eine echte Aufgabe aus seinem Job automatisiert.
Praktische Bewertung, Teilnahmebescheinigungen über eine Schulung zur KI-Kompetenz (Artikel 4 des EU AI Act), Start der laufenden Community. Vierteljährliche Auffrischungen geplant.
The training is delivered by specialists from the encorp.ai team. Their profiles are on the team page.
Co-Founder & CEO, encorp.ai
Author of the book “Survive AI” and a lecturer at SoftUni. Runs AI trainings for leadership teams and departments in banks, telecoms and media.
Kohorten in BG und EU — Fintech, Banking, E-Commerce, produzierendes Gewerbe, Dienstleistungsunternehmen.
Business media covering entrepreneurship, companies and the economy.
Technology company providing enterprise IT solutions and infrastructure.
Telecommunications operator providing connectivity and digital services.
Was bei KI-Training funktioniert, was nicht, und wie man die Adoption misst.

Marsh finds 78% of executives were offered AI training last year, against 20% of hourly workers. Adecco says training participation fell from 77% to 67%. Read together, the two surveys show who is left to learn AI alone.

Ireland's privacy regulator got X to stop training AI on data its agent collected about other people. For companies connecting agents to mailboxes and CRMs, the question is no longer only whether a vendor trains on your data, but what happens to the data your agent reads about clients, candidates and suppliers.

The Digital Omnibus delayed high-risk AI Act duties to 2027 and 2028, but bans, literacy, GPAI rules and most transparency duties already apply. Here is the calendar companies in Bulgaria and the EU should use now.
Next step
One of our experts will agree with you which teams to train, at what level and in what format, and send you a fixed-price proposal.
Or call us: +359 887 458 634
Or write to hi@encorp.ai