Xephor Solution is a cutting-edge Artificial General Intelligence (AGI) technology able to develop creatively and perform intellection tasks typical to humans. Xephor Solution can independently structure and analyze huge volumes of data or just few data at high speed and provide creative outputs. It is the only AI-technology worldwide which is 100% European origin and IP and the only working AGI worldwide.


  1. Intelligent systems:
    Ruled-based software: consisting of a database, a certain number of rules and a control system, which picks the suitable rules for the current problem case. Base of expert systems (if-else-sentences ->cognitions through analogies). Computer isn´t able to learn autonomously.

  2. Machine learning:
    Logic reasoning (IBM Watson): Rules are extended through logical terms and therewith the database is enlarged. Hidden-Markov-models (HMM) e.g. Siri: Relations are produced through statistics (probabilistic induction). Bayesian networks: Calculation of complex probabilities of former known results for further results.

  3. Deep learning:
    Inductive learning: Use of rules of generalization and specialization; valuable training data must be available. Reinforcement learning: Through trial and error worthwhile actions are found: trying -> response -> optimal strategy! Neural nets are working with transition probabilities (e.g. Google Deepmind): Artificial neurons with different connections and weights which are changing while learning.

  4. Machine thinking (Xephor Solutions) AGI:
    Network of neural nets. Replica of the human brain in an anatomical and functional way. Formation of hypothesis through neuro simulations with stochastic learning. New ideas („creativity“) and the ability to act to unforeseen situations are produced.


Training of the AGI-Brain

The software works in three steps to evolve and fulfil its purpose:

In the first step, the real user (employee) acts as a ‘teacher’ for the software to learn its initial tasks. The real user is communicating with the software embedded virtual user (artificial brain) to feed it the required information on tasks that need to be performed. This includes both delivering commands to the software as well as the software ‘observing’ the actions the real user performs on the computer.

In a practical sense, this means that the software can be considered to take the role of a new unexperienced employee, who needs to be taught their role and responsibilities and the tasks they will need to perform. In the case of Xephor Solution, the virtual user not only gets input from the real user, it is also observing what the real user is doing on their computer, meaning the software learns through the real user’s behavior.

In the second step, after the virtual user has learned from the real user, it can be creative in the same way as human beings are, and becomes autonomous in the tasks it learnt. The virtual user can improve its performance and eliminate any mistakes the real user made, based on its ‘thinking’ process. During this process the system sends random tensors into the neural networks, where the tasks it has been trained in are memorized. With that, the various output data are created:

With that, the brain looks for the optimal strategy to solve occurring problems just like a human being would do it. Through these simulations it can generate new output.

In the third step, after a period of learning (depending on the complexity of the tasks) the virtual user can work completely autonomously, which means that it is able to assess the environment and decide which action to execute and how to do it to achieve optimal results. Due to its ability to be creative it can react to every kind of new situation in real-time.

The whole training and interaction with the system is in natural language – neither technical know-how nor programming skills are necessary!


Market & Model of Cooperation

Our market is every sector, where you need creativity, highly skilled workforce and/or results, which should constantly improve.

We are working with cooperation partners in several areas:









Extract of areas we have already worked in:







Isabell Kunst

CEO & Co-Founder

Konstantin Oppl
Konstantin Oppl

CTO & Founder

Ralph Tippmann

Business Development/Investor Relations

Georg Jobst

Software Engineer

Friedrich Urbanek

Project Management & Co-Founder

Fedor Sapronov

Advisor to the Board


You want to get to know more about the most advanced AGI worldwide – just connect with us here!

Get in contact