Mere technological promise no longer, artificial intelligence has become a day-to-day business tool for Hungarian small and medium-sized enterprises (SMEs), according to presentations at the Portfolio "Lendületben a Hazai Vállalatok 2026" conference. Speakers gave practical examples of AI handling invoice processing, customer service, translation, international expansion, production planning and management decision support.
Examples of automation and operational tasks
Tamás Murvai, founder of AI Path Pro and an automation and AI specialist, argued that AI’s value for SMEs lies in automatable processes. In a live demo he showed a system that, given a natural language instruction, scanned an email inbox for invoices received during a given period, normalized varying names, filtered duplicates, renamed files and packaged and forwarded them to the accountant. Murvai used this example to illustrate that AI today is increasingly an operational colleague rather than only a digital assistant.
He also cautioned that not every task should be handed to a model automatically. For instance, forwarding invoices to an accountant is a controllable process, whereas automatically replying to customer emails should not be left to a model without supervision. Murvai advised that nothing be automated without testing it at least fifty times.
Management reporting and decision support
Murvai presented a hypothetical company case in which AI aggregated data from multiple sources — marketing, sales, customer and financial data — and produced management analysis, dashboards, presentations and verifiable Excel files. In that example the AI concluded the company’s problem was not a lack of leads but poor conversion, customer retention issues and product-page shortcomings. As he put it: “AI is a tool. The value is in the automation.”
Business outcomes and international expansion
Participants in the conference’s closing panel reported tangible business results. Norbert Pörzse, founder of the Vágyaim.hu online store, said that previously the firm relied on translation agencies, a slow and costly process for a catalog exceeding ten thousand products. After adopting a ChatGPT-based translation solution, they opened stores in 27 countries over three years; today 35 percent of their revenue — more than 1 billion forints — comes from abroad, a result Pörzse said would have been impossible without ChatGPT. He added that translating a full product database used to cost several million forints, whereas now it is affordable at a fraction of that price.
Legal risks and copyright
Dr. Ádám Miklós Sulyok, head of the Copyright Department at the National Intellectual Property Office (Szellemi Tulajdon Nemzeti Hivatala), warned about legal risks. Companies should be particularly cautious when uploading protected content, intellectual property or trade secrets into AI systems: “If we have such protections, we should think twice before putting them in.” He stressed that an author or inventor must still be a human, although some AI-generated elements, such as logos, can in certain cases receive trademark protection.
Production and workforce effects
Péter Galajda, managing owner of Mayer Szörp, described AI’s role on the production side as primarily data collection and preparing for later optimization. They measure every production batch and collect the data because those measurements will help AI-driven optimization. He also emphasized human limits to efficiency gains: it is important that colleagues survive the efficiency drive and come to work in good spirits.
Anikó Kaszás, CEO of Solvelectric Technologies Ltd., noted that automation has long been present in the electronics industry, so AI’s novelty there is more in analysis and eliminating time-consuming tasks than on the assembly line. Gergely Lakatos, vice-president of JEUNE, argued that time freed by AI should be deliberately reallocated to creative work; responsible decision-making remains with company leaders.
Measurement, controlled rollout and responsibility
A central takeaway from the panel was that AI’s return on investment is not always immediate: first it saves time, later it can reduce costs and ultimately open new revenue streams. Achieving these benefits requires clear goals, measurement and controlled implementation. Participants emphasized that AI delivers real business value only when companies measure outcomes, introduce automation in a controlled manner, and maintain human responsibility and legal awareness.
