Skip to main content

AI-Driven transformation: Challenges and Opportunities for SMEs and Public Organisations

Damir Medved for PSTW

Published on: 05/05/2025 News

This opinion piece is authored by:

Damir Medved

Damir Medved

Director - EDIH Adria

Damir Medved is the director of EDIH Adria (European Digital Innovation HUB), whose main goal is to enable small entrepreneurs and local self-government units to speed up the digitisation process of their organisations, speed up the innovation process, the opportunity to test ideas and solutions before starting the investment process and other related activities. He has more than 35 years of experience in managing complex transformation projects, and he is one of the initiators of the implementation of the Competence Centre for Smart Cities in Rijeka.

He is actively participating in several civil society associations with a goal to promote and preserve local heritage, ecology, citizen energy and the promotion of new technologies in these areas.

Contact details

  • Email address: damir.medved@uniri.hr
divider

Introduction

Global crises frequently act as catalysts for profound social and economic transformation, with the rapid advancement of Artificial Intelligence (AI) technologies playing a key role in driving these changes. AI has the potential to revolutionise workplace processes, expedite the adoption of digital collaboration tools, and ultimately broaden the use of AI technologies in public administration and business operations. Contrary to initial fears, many organisations have found that the application of AI technologies increase customer (citizen) experience, productivity and reduce operational costs even in small environments.

Advances in AI can help solve the chronic problem of labour shortages through the transformation of office work, enabling the automation of routine tasks such as communicating with citizens, facilitating the introduction of advanced analytical technologies, and improving the processing of large data sets. AI-supported public administration is no longer science fiction, but an emerging reality also in the Republic of Croatia. 

This article shows how AI affects the organisation of work in small and medium-sized public service organisations (PSOs) and enterprises (SMEs), how to enable their digital transformation, and finally, what challenges in the application of AI we will face in the coming period – all based on my personal experience as director of the EDIH ADRIA project.

The EDIH Adria project and the Digital Maturity Assessment in Croatia

The services provided within the EDIH Adria project are focused on the application of AI and high-performance computing primarily for smaller organisations. Specific areas of interest are public administration services, optimisation of traffic and energy flows, and the development of large data sets. The process typically begins with specialised bootcamps where users are given the opportunity to define a business problem or idea, and a digital maturity assessment (DMA) is conducted. Selected users are directed for testing before investing (TBI), where the quality of data, the definition of business processes and the technological maturity of the proposed solution are verified on a practical example. Rather than delivering ready-made solutions, EDIH Adria provides technology demonstrators – known as AI microservices – which serve as building blocks to help define and shape final solutions. The support journey concludes with the creation of a Digitalisation Action Plan (DAP), offering concrete guidelines for drafting tender documentation aimed at acquiring a full-scale production solution. In addition, through its FCO service, EDIH Adria assists beneficiaries in assessing their financial capacity, exploring EU co-financing opportunities, and connecting with relevant financial instruments.

DMA implemented in 40 SMEs and 44 PSO organisations within the EDIH Adria project pointed to major shortcomings that are present in smaller Croatian companies and local self-government units in the application of AI and advanced technologies. The rationale behind the DMA is that measuring the digital maturity status of an enterprise and its evolution over time (e.g. comparing its status before and after the EDIH intervention) will provide an indication of the effectiveness of EDIH’s intervention and, at an aggregated EU level, of the effectiveness of the EDIH policy initiative.

Challenges and opportunities for AI adoption in smaller organisations

Only 24.84% of Croatian organisations that participated in DMA is adopting AI in any form, and in data security and interoperability just 40.89% and 36.80%, respectively is implementing advanced technologies. The emergence of AI and digital technologies is fundamentally changing the dynamics of the workforce, presenting opportunities and challenges that organisations in Croatia are relatively struggling to cope with. Key issues include a general lack of documentation of business processes, poor understanding of the context of quality and transparency of data collection, and a general high level of distrust in the application of AI technologies to improve business processes. There is a serious shortage of trained AI consultants, as well as a lack of adequate training courses for employees that would be appropriate to their current level of knowledge. Given that the main labour shortage is in “daily operations” (driven by unfortunately largely unnecessary and excessive administrative processes), I can see two key problems:

  1. Lack of required skills: As AI automates routine and repetitive tasks, the demand for employees with advanced, specialised skills is increasing. This change requires a commitment to continuous learning and professional development. Organisations need to invest in training programs that equip employees with the necessary competencies to work with AI systems, such as process optimisation, data analysis, machine learning, and digital literacy. Addressing skills gaps not only improves employees' abilities but also ensures that the workforce remains adaptable to evolving technological trends. Yet, a fundamental problem here is the shortage of suitable education programs.

  2. Digital transition-related issues: Automation and AI have the potential to replace jobs in certain roles, leading to anxiety and uncertainty among employees. To alleviate these concerns, organisations should develop comprehensive reskilling and upskilling initiatives that provide pathways for employees to transition into new roles created by technological advancements. This proactive approach demonstrates a commitment to employee well-being and helps preserve organisational knowledge and expertise. However, this is not always the case, especially in smaller rural areas.

On the other side, the positive fact is that, from what results in the activities of the EDIH Adria, most owners and management (mayors in public administration, especially in smaller communities) are aware of the needs of digital transformation, informing employees, as well as the importance of education and are ready to finance them.

As seen in EDIH ADRIA project, the rapid advancement of AI and digital technologies has opened the potential for a serious transformation of jobs even in small public organisations or SMEs, significantly improving opportunities for efficiency, innovation and growth. However, this technological revolution also raises numerous questions and doubts that are often not adequately addressed. Among the most pressing issues are those related to bias and discrimination, privacy and protection, safety and (mental) health, legal and regulatory frameworks, and the dynamics of the evolving workforce. 

Addressing these areas is key to supporting trust, ensuring compliance, and fostering a resilient and adaptable organisational culture – and this is recognised by EDIH ADRIA users. AI and digital technologies can have a major impact on fundamental human rights, especially when it comes to equality and non-discrimination. Inadequately used AI tools (especially in public administration) create the risk that AI could further contribute to discrimination. The algorithms used to inform citizens, process and evaluate their requests, manage processes and assess impact can sometimes introduce biases that subtly put people at a disadvantage based on characteristics such as gender, age or ethnicity. It is in this context that we have witnessed concerns among our users of AI technologies.

However, the important fact is that AI itself is not and should not be biased. When professionally designed and used, AI can help reduce prejudice and discrimination, making decisions fairer. An excellent example of this comes from the small town of Kastav where an AI-based tool that processes citizens' requests makes it more transparent and efficient than human evaluators, as illustrated in the box below. 

 

AI Use Case: City of Kastav

The City of Kastav serves as a local public authority committed to fostering education, environmental protection, and cultural initiatives, including support for renowned events like the Kastav Summer Festival.

During execution of EDIH ADRIA TBI service, activities were conducted with the aim of digitising the process of handling application documentation for co-financing tenders conducted by the City of Kastav. The City of Kastav has a need for partial automation of the process of receiving and processing tender documentation to facilitate the implementation of tenders in various fields (culture, sports, business incubators, etc.).

The subject of the service was the testing of a large language model (LLM system), initially intended to fulfil the following functionalities:

  • Submission of tender documentation in digital form,

  • Automatic completion of the application form with applicant data,

  • Creation of evaluation proposals based on the provided criteria.

As part of the TBI service, examples of project proposals and accompanying documentation submitted in response to a previously conducted public call for project proposals were analysed. A prototype AI model was developed, capable of extracting data from tender documentation, generating a summary of the project proposal, evaluating the proposal based on the provided criteria, assigning scores for each defined criterion, and outputting the evaluation results in a table format that can be exported to Excel and stored on a computer.

As result, incoming project proposals can be ranked accurately, enabling commission representatives to create a scoring ranking list of the received project proposals quickly and transparently and to easily access the most important and relevant information from each submission.

 

Conclusion

As the results of the EDIH Adria project have shown so far, AI stands out as a transformative force that can reshape public administration or SMEs and redefine their operational paradigms. 

For organisations that want to harness the full potential of AI for digital transformation, a gradual, strategic, and multidimensional approach is crucial. This includes investing in employee training, promoting a culture of transparency, prioritising ethical standards, and engaging with all stakeholders in the transformation process. By addressing these key areas, even small communities can effectively AI technologies, and ensure that innovation is driven while maintaining trust and accountability.

divider

We want to hear from you!

Please use the comment section below to share your insights and start an insightful discussion.

Login or create an account to comment.