Scientific processes could become faster and research more efficient thanks to the latest initiative from HUN-REN HQ. From October, members of the research network will have access to unique AI services, with dedicated AI specialists available to help them choose and use the most suitable AI-based solutions.
Launched this year, the initiative aims to provide researchers in the network with access to a wide range of AI services, enabling them to make their research more efficient, accelerate research processes, or even target entirely new research topics.
The AI4Science programme, HUN-REN HQ's AI service portfolio, is based on the research needs identified and articulated within the network. It is available to all researchers and helps them achieve even more cutting-edge results. Through the programme, HUN-REN encourages researchers to integrate AI tools into their work and to take full advantage of the support AI offers across all fields of science.
"Our global competitiveness is also supported by adopting and utilising the latest scientific and research technologies and ideas, as well as by researching and disseminating them," emphasised Balázs Gulyás, President of HUN-REN, at the AI4Science programme launch event.
"Our goal is to reduce the time our researchers spend on administrative tasks and increase the time they can dedicate to research activities. To support this, we are providing AI-enabled technologies and platforms. In addition to offering the appropriate AI infrastructure, we are continuously gathering best practices from around the world, whether in training or educational programmes," highlighted Roland Jakab, CEO of HUN-REN.
The cornerstone of the AI4S programme is the pan-institutional AI Ambassadors' Network, comprising 25 members, who will run AI workshops and training sessions and help identify and exploit the benefits of using AI. Additionally, researchers will be able to access professional help from them or, through them, from a central team of experts, both for the validation of research ideas and for infrastructure support.
Other areas of AI support provided by HUN-REN HQ:
- AI education and inspiration: Researchers can participate in both basic and specialised training courses.
- Personalised AI support: Researchers have access to the AI expert team at HUN-REN HQ, an external team of specialists, and an AI research partner-matching service to help find tailored AI solutions.
- Available AI technologies: To protect sensitive research data, researchers can use HUN-REN’s proprietary AI framework and high-performance computing infrastructure.
The implementation of AI4S and the work of the AI Ambassadors are supported by three AI experts: Gergely Szabó (AI research idea validation and research design), Dániel Szittya (AI applications), and Gergely Kóczán (AI infrastructure). If further assistance is needed, an external team of specialists from Neuron Solutions is available to HUN-REN researchers or ambassadors for specific inquiries related to AI.
AI technologies already available to researchers
"AI technologies are already available at HUN-REN," as explained at the launch event. "Within our own AI framework, researchers can use the HUN-REN Cloud, a closed, cloud-based platform operated by SZTAKI and the Wigner Research Centre for Physics. The advantage is that sensitive data remains protected and is stored exclusively on servers located in Hungary."
The HUN-REN Cloud hosts the GenAI4Science service developed by SZTAKI to support generative AI tasks. It enables researchers to harness the potential of large language models (LLMs) while ensuring secure data storage in line with data management policies. Thanks to this new service, they can use public, open-source models in an easy-to-use chat mode, even incorporating their own data, documents, and other sources.
The Komondor computer is also available to HUN-REN members. Operated by the Governmental Information-Technology Development Agency (KIFÜ), it is the most powerful supercomputer in Hungary, launched in January 2023 and equipped with over 200 GPU cards.
- How AI can improve researchers' work
-
At the event, the benefits of AI were illustrated through a practical example. When breeding a plant that can withstand climate change, AI can both select individuals with better genetic traits more quickly and accelerate scientific processes. This allows researchers to spend more time on scientific activities rather than counting the number of new leaves day after day, as AI can do the work.