Logo webu
Are you interested in studying at Unicorn University? Sign up for an open day and find out more.

Unicorn AI Research Center

We are creating the future of artificial intelligence
The Unicorn AI Research Center (URC) is a research center operating under the Unicorn University, dedicated to research in the field of Artificial Intelligence (AI). For several years now, AI models have been experiencing a major boom, which we at the URC want to participate in. Behind the current wave of attention are large language models, in particular the popular ChatGPT tool, which offers many new possibilities but also reveals potential risks to keep in mind when developing AI.
URC focuses on research across a wide range of artificial intelligence tools, be it image recognition, text analysis and classification, cybersecurity, language models or statistical models.
The purpose of the URC is not only research in the field of AI, but also exploring the integration of existing AI-based tools within products for various market sectors, e.g. education, banking, insurance, energy, transport, engineering, etc.
We are currently working on the integration of large language models in the development of the uuBusinessChat product, which includes a superstructure used to make the language model and its modification for specific use cases. We also focus on the development of AI-based prediction models in the energy sector, where machine learning tools also have great potential.

Projects using AI at Unicorn University

Social and Motivational Factors of Study and Their Impact on Learning Outcomes in Tertiary Education with a Focus on Technical Fields.
Use of neural networks for realtime prediction of variable baseline in power equipment at HV, LV voltage levels

Unicorn system develops its own AI - Project uuBusinessChat

We have been working intensively on artificial intelligence applications for several years within Unicorn Systems, where we develop products that can work with artificial intelligence. Our great advantage is a detailed knowledge of the environment and enteprise/IT architecture of our clients, so we can understand this area comprehensively and think through all its implications.

The use of LLM models in the enterprise segment means in particular tokenization of custom content and subsequent work with it in combination with publicly available language models. This area is addressed by the uuBusinessChat product, which has the following features:
We have a content management system (uuBooKit) that can store content in a way that is optimal for subsequent tokenization of this content. We can also transform data from other
existing sources on the customer side, whether structured or unstructured. In our CMS, the content is also associated with the access rights of the deVnated roles, so that when querying
only the information to which it has the necessary permissions is provided.
We can embed the required content - embedding is a special data representation format that can be used in machine learning models. When embedding, we use our own algorithms for preprocessing and
markup of content. Embedding can be done on a regular basis, so the updated content gets into the language model quickly. The content is fully under the control of the client, it can be stored in a private or public cloud.
We were one of the first to get access in Microsoft Azure Open AI Service - this is a service that provides an API to ChatGPT's LLM models that are usable in this way in custom applications.
The uuBusinessChat custom app combines the above technologies and allows you to chat about private content. We use special algorithms and hardware to optimize search and unlike public
ChatGPT, we also provide source content to the search result. uuBusinessChat is a cloud-based microservice application that can be used independently or embedded in other applications.

Ambassadors

Karel Šafr


Contact: karel.safr@unicornuniversity.net

Vít Stinka


Contact: vit.stinka@unicorn.com
document_check.svg
We use cookies on this website to ensure its functionality and to personalise ads, solely with your consent and in accordance with our Cookies Policy.

By clicking on the "Accept cookies" button, you consent to the use of selected cookies and agree to the transfer of behavioural data for the display of targeted advertising on social and advertising networks. You can choose which information you want to share with us by clicking on the Cookie settings button.