MB Kaip reikiant

Innovation · AIOrchestra · 2026-05-31

Lithuanian start-up beat Fujitsu by three days with its innovation.

Fujitsu made it loud. A Lithuanian start-up had already moved from public market signal to working codebase and standalone domain.

Skaityti lietuviškai

On May 22, 2026, a Lithuanian start-up under the MB Kaip reikiant umbrella introduced AIOrchestra, a self-evolving multi-AI agent system for real business work: analytics, research, marketing, copywriting, CRM, testing, routines, and learning cycles.

On May 23, Git history shows AIOrchestra being separated into its own working codebase with Telegram-based intake and control already present. On May 24, the dedicated aiorchestra.eu domain was registered.

Three days later, on May 25, Fujitsu publicly announced its self-evolving multi-AI agent direction.

Fujitsu did it at corporate scale: with an official announcement, international PR, and a clearly articulated technology direction.

The Lithuanian start-up did it differently: quietly, practically, and from Lithuania, by first showing a working direction rather than a loud promise.

We respect Fujitsu's contribution to AI technology development and business innovation. We also understand that corporations of that size do not have to notice small builders like us. But scale and chronology are not the same thing. Fujitsu made this direction loud on May 25. A Lithuanian start-up had already brought AIOrchestra to the market under MB Kaip reikiant, moved it into a working codebase, and registered the standalone domain before that.

They made it loud. We had already made it visible, workable, and standalone.

What matters here?

Not that a small Lithuanian company speaks louder than an international corporation. It does not.

What matters is that the direction now being clearly articulated by major technology players had already started to be realized by a Lithuanian start-up earlier.

AIOrchestra is not another chatbot. It is a managed AI employee team where agents have roles, levels, work rhythm, tests, approval gates, cost control, and learning mechanics.

In other words, AI does not only answer questions. AI works, coordinates, leaves evidence, learns from mistakes, and helps humans make better decisions.

How is this different from Fujitsu?

Fujitsu naturally speaks to enterprise scale: large organizations, complex systems, regulated industries, and critical infrastructure.

AIOrchestra is broader in practical terms: the same logic can serve not only corporations, but also growing businesses, small teams, solo founders, or even beginning bloggers who need not another tool, but a working AI team.

That is why Fujitsu's announcement is not a starting signal for us.

It is market confirmation that the direction is right.

Lithuanian innovation does not always start with noise

Small teams often do not have a large PR machine. They do not have global campaigns prepared for months. They often just build, connect, test, fix, and release.

That is how AIOrchestra appeared.

It was born not as a loud theory, but as a working system: with agents, Workroom, BackOffice logic, Hacker tests, Staff Officer coordination, daemon routines, and a real need to automate business processes so the human does not have to push every small step manually.

Fujitsu announced a similar direction loudly.

A Lithuanian start-up started realizing it earlier: public page on May 22, working codebase on May 23, standalone domain on May 24.

Even if only by three days.

In innovation, three days are still chronology.

Read more

A more detailed English AIOrchestra publication about the relationship between AIOrchestra and Fujitsu's direction is available in the AIOrchestra blog: AIOrchestra was already moving before Fujitsu made self-evolving AI agents loud.

Sources: