AI chatbot and AI agent development

Mesterséges Intelligencia Központ Kft. is a software development company based in Budapest, Hungary. We build AI chatbots and AI agents that answer from the company's own documents (in Hungarian too) and, connected to the company's systems, also carry out tasks.

In short: the chatbot answers from the company's own documents, in Hungarian too, citing the source; if the answer is not in the material, it says so and hands the question to a member of staff. An AI agent also carries out steps in the company's systems, with access rights and approval before important steps.

Chatbot or AI agent?

Both can run on the same language model; the difference is in what they are allowed to do. Our article Chatbot or AI agent? covers the differences and the risks in detail.

AspectAI chatbotAI agent
Its jobanswers questionsreceives a goal and carries out steps
What it can accessthe knowledge basebesides the knowledge base, tools: CRM, calendar, email, order handling
Example“When will my order be delivered?”“Please record the complaint and send the customer a confirmation.”
Controlcited sources, logged conversationsaccess rights, approval before important steps, a log

Answers from the company's own documents, in Hungarian too

The chatbot works from the company's own material: FAQs, product descriptions, policies, manuals, the general terms and conditions. The method is called RAG (retrieval-augmented generation): for each question the system first finds the relevant passages, and the model answers from them, citing the source. If the answer is not in the material, we build the bot to say so and hand the question to a member of staff instead of making something up.

Today's large language models understand and write Hungarian well. Before going live, we test the bot on Hungarian questions with typos and everyday phrasing, and when a document changes, the knowledge base is updated too. Our article AI on the company's own documents (RAG), in Hungarian too explains the method in detail.

Where does a chatbot or agent help?

When an agent connects several systems, it is also process automation; we write about connecting language models and about data protection on the AI integration page.

A live example: the AI assistant on our homepage

Next to the game on our homepage, the AI assistant talks with visitors: it is our own live AI chatbot. What it does:

Ask it, for example, what the company does. You can read about our own projects on the Our work page.

Transparency: people must know an AI is answering

The European Union's regulation on artificial intelligence is the AI Act, officially Regulation (EU) 2024/1689. Under its Article 50(1), AI systems intended to interact directly with natural persons must be designed so that the people concerned are informed that they are interacting with an AI system, unless this is obvious from the circumstances. This rule has applied since 2 August 2026. Source: the text of the AI Act on EUR-Lex.

That is why a chatbot we build states in its very first message that an AI assistant is answering. If the bot also handles personal data, the GDPR applies as well. This is information, not legal advice.

How is it built?

  1. Free first consultation: who the bot talks to, about what, and what it is allowed to do.
  2. Knowledge base and access rights: which documents it works from, which systems it can see and what it may change.
  3. Size measurement and offer: we measure the size of the development in COSMIC function points, and the offer and the schedule are based on it (see Pricing).
  4. Development in stages: tested on real questions, with regular demos.
  5. Handover and operation: the contract governs source code and usage rights; after handover we also keep the knowledge base up to date and monitor how the bot performs.

Frequently asked questions

What is the difference between a chatbot and an AI agent?

A chatbot answers questions from the knowledge base it is given. An AI agent also uses tools: it looks things up, records, books and sends emails, so it carries out steps towards a goal. With an agent, designing the access rights and the approval points is therefore an important part of the work.

Does it answer well in Hungarian?

Yes. Today's large language models understand and write Hungarian well, and they work well with Hungarian documents too. Before going live, we test the bot on real Hungarian questions; the AI assistant on our homepage also talks in Hungarian and 13 other languages.

How does it know the company's data?

From what the company gives it: the uploaded documents and the systems it is granted access to. Access can be narrowed so the bot sees only what its task requires; it can cite the source next to its answers, and conversations can be logged.

Do we have to tell users that an AI is answering?

As a rule, yes: under Article 50(1) of the AI Act, people must be informed that they are interacting with an AI system unless this is obvious, and this rule has applied since 2 August 2026. A chatbot we build says so in its first message. This is information, not legal advice. If you are planning a chatbot or an agent, write to us using the details on the Contact page.