Visit to COINIS: Exploring Opportunities for HPC and AI Collaboration

On 8 April 2025, the NCC Montenegro team—Tomo Popović, Ivan Jovović, and Elvis Taruh—visited COINIS, a leading Montenegrin digital marketing and ad tech company, to discuss potential collaboration in high-performance computing (HPC) and AI. We were warmly welcomed by our hosts, Ms. Tamara Pavlović and Ms. Aida Ismailisufi.

Visit to Coins to discuss HPC opportunities

The NCC team introduced the EUROCC2/EUROCC4SEE project and discussed how COINIS could leverage HPC resources for advanced data analytics, AI-driven personalization, and generative AI applications. Opportunities through EuroHPC calls, including FFPLUS, were presented, alongside examples of successful industry use cases.

Further collaboration was discussed, including POC project support and joint capacity-building activities. A follow-up is planned to define a concrete use case and collaboration with regards to a possible application process.

There was around 10 people from the company interested in the presentation

Online Course on Prompt Engineering Starts on 7 April

We have just enrolled over 25 people interested for the Prompt Engineering course. The course is implemented as a self-paced online course via LMS. We mainly enrolled students at BSc and MSc level, but there are few people coming from the industry sector (mainly IT). Most of the students are from Montenegro, but we have a couple of people coming from Bosnia and Serbia. Our target is that all of the enrollees finalize their work by 1 May.

The course starts on 7 April
This is a self-paced online course implemendet via LMS platform

Empowering the Future with AI: Workshop on Chatbots and LLMs at NCC Montenegro

As part of the ongoing H2020-MSCA TRACEWINDU initiative to foster digital innovation and transparency across value chains, a hands-on workshop on chatbot development using Large Language Models (LLMs) and the Azure platform was held at the National Competence Center for High-Performance Computing (NCC Montenegro), hosted by the University of Donja Gorica (UDG).

This workshop was implemented as a cross-collaboration of TRACEWINDU and EUROCC2/EUROCC4SEE ptojects

The workshop brought together a vibrant group of PhD, Master’s, and Bachelor students, alongside AI and HPC researchers and mentors from UDG—key TRACEWINDU and EUROCC4SEE project partners—highlighting the growing interest and expertise in cutting-edge AI technologies across disciplines.

Leading the session were Ms. Daliborka Nedić and Ms. Senka Gajinov from DunavNET, an innovation-driven technology company and core partner in the TRACEWINDU consortium. Their expert lecture provided attendees with valuable insights into the rapidly evolving world of generative AI and practical guidance on building intelligent chatbot solutions using Microsoft’s Azure OpenAI services.

Around 20 people attended, mostly PhD, MSc and BSc researchers interested in AI and HPC

The workshop covered key concepts such as:

  • The fundamentals of Large Language Models (LLMs) and their applications in chatbot development
  • Integration of LLMs within the Azure AI ecosystem
  • Best practices in prompt engineering and knowledge grounding
  • Real-life use cases aligned with transparency, automation, and user engagement in value chains

Beyond chatbot development, the discussion extended to the integration of computer vision models—particularly those aimed at automating data collection in value chains—and the use of HPC resources to support their deployment. Participants explored ideas for combining conversational AI interfaces with real-time computer vision capabilities, opening up possibilities for smarter, more interactive systems across agri-food and other sectors.

This collaborative brainstorming set the stage for future joint initiatives, where chatbots could serve as intuitive front-ends to complex AI models running on high-performance computing infrastructures—bridging accessibility and computational power for impactful solutions. This workshop is part of a broader capacity-building effort within TRACEWINDU and EUROCC4SEE, designed to empower young researchers and professionals with the tools and knowledge needed to innovate responsibly in the age of AI TRACEWINDU continues to bridge digital technologies and real-world value chain transformation.

Ms Daliborka Nedic and Ms Senka Gajinov from DNET at UDG

Invitation to participate in a short HPC/AI training for business users

We invite your company to join our specialized training on High-Performance Computing (HPC) and Artificial Intelligence (AI), designed for business users. The training will take place on April 16, 2025, at 10:00 AM at the University of Donja Gorica (UDG).

Introduction to HPC is a short course that requires no prior technical knowledge. It is designed to provide beginners with fundamental insights into high-performance computing, addressing key questions: WHY and HOW to use supercomputers.

This course offers practical insights into how HPC and AI can enhance research, development, and innovation, optimize business processes, and transform business models across various industries.

HPC/AI training for business users

During the course (90 minutes + direct consultations upon request), participants will:

  • Gain an understanding of the technical and commercial aspects of HPC systems and applications
  • Explore the possibilities and benefits of HPC and AI through industry examples and success stories
  • Learn how to assess readiness for adopting HPC technologies
  • Learn how to access Europe’s top supercomputers (at no additional cost! 😊).

The course is tailored for managers, experts, and professionals who want to understand the potential of modern technologies and their applications in everyday business operations. It is ideal for those seeking advanced computing solutions (HPC) and intelligent algorithms (AI) to enhance their work and business.

Join us and discover how HPC and AI can become your competitive advantage!

📅 Date & Time: April 16, 2025, at 10:00 AM
📍 Venue: Amphitheater AS, University of Donja Gorica
📝 Registration: https://forms.gle/wHq31ioQaQbLjqMU9

📩 For more information:  Sanja.nikolic@udg.edu.me

Consultations on a new application for HPC resources with PAID Montenegro

Members of the NCC Montenegro team, Prof. Dr. Enis Kočan, Prof. Dr. Nikola Žarić and Assist. Prof. Luka Filipović visited on Friday, March 28, the company PAID Montenegro, which represents an example of efficient use of HPC resources, as well as an example of good cooperation with NCC Montenegro. Namely, as part of the development of its stock exchange trading model, PAID-T, the company received access to HPC resources on the LUMI supercomputer in April 2024. Prior to this application, the NCC Montenegro team had several meetings with representatives of the PAID Montenegro company, where they were presented with the services of the national HPC competence center offered within the EuroCC2/EuroCC4SEE projects, and were also provided with support in completing the application for the Development call. Even before the expiration of the one-year deadline, through work on the development of the PAID-T model, the company has spent the allocated 15,000 node-hours on partition C of the LUMI supercomputer, and is already preparing a Report on the results and experiences so far.

Meeting with Paid MNE

Since the work on the development of the PAID-T model is not completed, the PAID Montenegro company plans to re-apply to one of the calls for access to supercomputer resources in Europe. Continuing the good cooperation, the NCC Montenegro team first presented the current and future calls organized within the EuroHPC JU ecosystem. Through an analysis of the expected future needs for HPC resources of the PAID-T company, taking into account the dynamics of publishing calls for access to HPC resources by the end of 2025, as well as the resources made available to users through these calls, it was proposed that the PAID-T company re-submit an application to the same type of call (Development call). It was suggested that they also apply for resources on the LUMI-G partition this time. It was agreed to continue contacts with the NCC Montenegro team and provide support in writing a new application for HPC resources.