Tóth Tamás, lead expert at AI Workshop, reports that during a 27-day vacation he worked only 30 hours while AI agent teams he built completed tasks across his two companies that would have taken roughly 11 months with traditional organization.
Which companies were involved?
The experiment involved two of Tóth Tamás's own businesses: one focused on software development and the other providing fitness services. According to his account, both businesses benefited from integrating AI to accelerate processes and improve scalability.
How did the AI agents operate?
The approach did not rely on single prompts. Instead, AI "colleagues" were assigned distinct roles, specific tasks and goals, and rules for collaboration. These agents worked in a coordinated way, relying on each other's outputs to carry out management, research, planning and execution tasks.
What concrete tasks were completed?
The systems were used in software modernization, marketing, brand building, administration, reporting and financial process automation. One highlighted example: the full redesign of a 22-year-old system interface, which previously would have taken an entire summer, was completed within the first 10 days and is now in use. The AI also participated in brand identity and logo design, assembled newsletter campaigns, and prepared 30-day product launch strategies.
Organisational lessons
Tóth emphasizes moving beyond seeing AI only as a conversational partner; effective use requires delegating tasks and creating structured workflows. He argues AI is especially valuable for taking over routine, repetitive tasks, freeing human employees to focus on creative and strategic work.
Data security and subscriptions
Tóth warns about a common corporate risk: employees using private AI subscriptions for work. Data uploaded to private accounts can be used by AI providers for model development, whereas Business/Enterprise subscriptions typically guarantee data protection. He advises company leaders not to skimp on appropriate enterprise subscriptions and to establish clear internal rules for AI usage.
Training and dissemination
AI Workshop aims to provide practical AI skills to Hungarian business leaders. For the upcoming August training, participants will not receive theoretical lessons but will develop solutions tailored to their own company problems within a workshop format. The program teaches how to build agent teams that operate autonomously and communicate with one another in the background.
Conclusion
The experiment indicates that with proper planning and role design, AI agent teams can markedly speed up business tasks and reshape traditional operating models. At the same time, data protection and corporate governance are essential for the safe and sustainable adoption of the technology.
