Tokić Builds Its Own Digital Platforms – Developers Create Solutions for Real Users, Not Outsourced Projects

Digital transformation, according to Josip Kelava, Technology Director at Tokić Group, “is first and foremost a mindset—a way of thinking that starts with people.” Having joined the company six years ago, right at the peak of its transformation journey, Kelava emphasizes that real change begins with employees, departments, and the decisions they make every day.

“We’re a truly data-driven organization,” Kelava says. “Nearly every key decision is based on data.” Analytics has become a strategic pillar of the group, embedded across all departments—with more than 180 active reports underscoring the scale of that commitment.

The company established its first analytics team in 2019 within the Digital Transformation and Innovation Department, which later evolved into a standalone division split between data engineering and business intelligence. “One team handles data pipelines and validation, while the other focuses on visualization and executive reporting,” Kelava explains.

Tokić has been digitally transforming itself from new ERP systems, DWH solutions, and web shops, to the modernization of call centers. What are the key phases of that process, and where do you see the greatest challenges in their implementation? How did you manage to connect different systems into a unified whole that brings real value to customers and employees? How much have these solutions improved the end-user experience, and how much have they facilitated work within the organization itself?

That’s right, we have decided to invest in many areas, so this is a very intense period for the technology departments within Tokić Group, but that is exactly the magic of working at Tokić. Because we are much more than a retail store. Each of the systems we develop is a story in itself, but certainly the largest one is ERP, which is the engine of the company, the main (core) application to which all others are connected. The implementation of such an extensive and demanding system as Dynamics 365 Business Central can entail certain costs.

For the implementation itself, we chose the proven Microsoft methodology called “Sure Step,” which ensures successful completion, on time and within budget. It consists of six main phases: diagnostics, analysis, design, development, deployment, and production. After the diagnostic and analysis phases, the current state is understood, then all new user requirements are consolidated, and based on that, the system is designed to integrate everything into one whole.

All of this is done to improve the user experience, speed up integration with partners, close the “technological debt,” and achieve operational excellence – and we start production in December 2025. We implement other projects using a similar methodology, and they are still in development, which is why I say that these are currently exciting times for us in the IT world of Tokić Group.

Digital transformation is often mentioned, but rarely clearly defined. How did the beginning of that journey look for you, and what would you single out as a key milestone along the way? To what extent has the transformation affected work culture and organizational processes? Did employees accept the changes, or was it necessary to invest additionally in education and adaptation? Which parts of the transformation would you rate as the most successful, and which required the most patience and resources?

I would definitely define digital transformation as a “mindset,” more precisely, a way of thinking, and by that I am referring to people. It is the transformation of a company through its employees, departments, and the decisions that are made. The founding of the Department of Digital Transformation in 2019 was the official moment, but the “mindset” existed even earlier, so the founding of the Department was merely the result. Tokić was the first to give a chance to the then-startup Gideon Brothers to test their autonomous robots in our warehouse. We introduced RPA technology into the company, founded a data analytics department, and applied the Kaizen principle of work – all this shows that we foster innovation, give a chance to new technologies, and invest in the future.

Recognition also came from the London Stock Exchange through the ELITE program, as “Future Shapers,” placing us among the 50 most innovative companies in Europe. That was a turning point that greatly changed the view of processes and the way of thinking, and the right questions began to be asked, the so-called 5W (who, what, where, when, why). Some employees at first showed resistance, out of natural fear (that results might be wrong, that something might not be delivered on time, etc.), and there were also unjustified fears regarding the development of RPA (automated processes) – such as that robots would replace them. Colleagues no longer had to do boring and repetitive tasks; they had more time for work that brings added value to the company, and absolutely no one lost their job; on the contrary, we only kept hiring and expanding to this day. We worked hard to educate our colleagues, to show them by example that this is just an aid and nothing else – we were persistent and we succeeded.

My personal view is that we successfully and quickly completed reporting and automated the operational parts of the processes, but what took the most time and resources was the digitalization project (introducing a digital archive, digital signature, and generally removing physical papers from the company). It took a long time working in one way, so colleagues were tense about throwing a physical paper into the trash, and so on. Currently, digitalization is deeply woven into the DNA of the company, and colleagues themselves come up with suggestions for improvements and make maximum use of the benefits of digital transformation.

Automation through RPA technology opened the door to our own innovations. Can you describe how the transition from ready-made tools to the development of your own solutions looked? How has such an approach proven more efficient compared to using external solutions? What has it brought you in terms of flexibility and business scalability? Are there examples where the development of internal solutions has brought unexpected advantages?

It has absolutely proven to be a good approach to develop our own RPA solutions. We started by hiring one specialist who “broke the ice,” and soon another team member joined through our internal academy, through which we scout talent. When you have your own in-house development, you stand at the source and dictate the pace yourself, assign priorities, and make decisions on the go. That made us more efficient, faster, and cheaper. Large systems and solutions require implementation, training, and licenses – and you still need an RPA specialist for development, input, and maintenance. We got all of that in our own specialists who can program by themselves, and it greatly helps to have a team member who acts as a bridge between business and technology – understanding both sides and translating them into mutually understandable language.

On the other hand, that also brings responsibility for system stability and maintenance, as well as potential risk in case of employee departure. We went through all of that and still wouldn’t change the path, because the results are there – we are approaching the number of 300 developed RPA processes, which annually save over €350,000 and perform around 12 million tasks.

What we could not have expected was the fact that we entered the field of artificial intelligence long before ChatGPT was launched, when the hype around AI began. Through research and development, we realized that RPA and AI work in symbiosis, that they are compatible and complementary – to put it simply, RPA executes while AI thinks.

Artificial intelligence and predictive tools are becoming increasingly important in business. How do you use AI, for example, through Lokad or similar systems, and what results do you achieve? Can AI, in your case, completely replace human judgment, or is it a combination that gives the best results? Which processes does AI improve the most – demand forecasting, stock optimization, or customer support? How much did the implementation of AI solutions require adjustments to internal processes and teams?

As we scratched the surface of artificial intelligence through the development of RPA, we sensed an opportunity for new, significant savings in operations. We took thousands of invoices and labeled them to create a data set needed for the approval system, digital archive, and document flow process within the company. That’s how we came to our first application, SimplyDoc, which, using artificial intelligence, reads and inputs about 100,000 metadata entries from various documents into the system – all of which were previously entered manually. Then came the opportunity to improve availability and optimize the stock of car parts in the supply chain, for which we engaged the renowned French company Lokad, which also works with Boeing, specializing in the supply chain.

We use AI and ML (machine learning) in ordering goods, where we learn the behavior of each supplier and the delivery discipline of their assortment. We have significantly increased our availability (95% constantly at all locations in Slovenia and Croatia), optimized inventory turnover for fast-moving items, and all in line with the payment terms of individual suppliers. We were required to maintain data hygiene, change our way of working, and accept new technologies that we needed to monitor at the beginning. New models and applications are arriving very quickly, and possibilities are more accessible than ever. It is a very exciting period, and I am happy about the new initiatives we are currently working on.

It is important to emphasize that none of this would work without colleagues who are at the beginning and end of the process – everything in between is done by AI. Therefore, the theory that artificial intelligence will replace us falls apart.

Analytics and the work of the data team have become the foundation of strategic decisions. How is your team structured, and how are the results of their work embedded into everyday business? Can you give a concrete example of a decision or project that came entirely from data analysis? How do you ensure the quality and accuracy of the data on which decisions are based? How are analysis results communicated to management and operational teams?

We are a “data-driven” company – all important decisions are based on data, as much as possible, of course. Analytics is a strategically important segment that runs through all departments of the group – the very number of about 180 reports shows how much we invest in it. Already in 2019, we founded an analytics team within the Department of Digital Transformation and Innovation, which eventually became a separate department. Now, internally, we divide them into data engineering and business reporting. One team prepares and brings in data, checks accuracy, and stores it for end use – i.e., for the visualization layer (reports), which is handled by another set of colleagues closer to management and the board.

This has proven necessary in the last few years because business knowledge and data science are interwoven. Thus, last year we launched a project to implement a new data warehouse (DWH), with the goal of faster and better reporting for management and the board – but also as a foundation for AI, which requires a large, structured, and fast database of accurate data. Various decisions are made based on our reports; just a few examples are the choice of locations for new branches (based on calculations of potential and vehicle fleet), expansion of assortment, and targeted field visits.

Tracking sales at the level of item, customer, program, margin share, web sales and customer behavior, warehouse and transport efficiency, and financial controlling – these are just some of the many opportunities for decision-making based on analyses that we deliver through a tool available to everyone, even on mobile phones.

Technology departments are becoming a key driver of development in every modern company. What are your plans in terms of expanding services and further strengthening these departments? Where do you see the greatest growth potential – in new technologies, further process optimization, or internationalization of services? How much do you rely on partnerships and collaborations, and how much on internal development? How do you plan to attract and retain experts in an increasingly demanding technological environment?

The primary goal and purpose of our teams is to support the operational part of the business – we are engaged in the sale of auto parts. Of course, there are additional values for our customers in the digital world, such as online ordering systems, mobile applications, loyalty programs, technical vehicle information, and many other digital touchpoints.

We constantly work on new services and platforms, and we currently see an opportunity in providing remote technical support to our customers through a combination of remote vehicle coding devices and communication software. Since modern vehicles are computers on wheels (around 300 million lines of code), we also have to keep up with the vehicle fleet and provide our customers with accessible tools.

All of this is done in cooperation with partners and with one eye on foreign markets, where we see enormous potential for the group’s growth. Internally, we are focused on the mentioned projects, and the plan for the future is to be faster, stronger, and better in applying artificial intelligence, which can bring us outstanding results.

We are active in the field and at tech conferences, members of the global digital commission within the most influential trade IAM organization, ATR International, and from the very beginning, we have been part of the Croatian Artificial Intelligence Association (CroAI), nurturing a culture of innovation. This is recognized and appreciated, so we have no problem finding new talent.

Moreover, when people realize that Tokić is not just about car parts but that we internally and directly develop technologies for our own needs, they reach out to themselves with the desire to work at Tokić. We have an environment where developers and analysts do not work for third parties; they are not outsourced for someone else’s business, they barely know, but operate directly and internally in a business environment where our technological product will actually be used.

That’s why, as developers, we can enjoy our work and give our best while sitting on real data, with precise instructions and freedom in our work. That is why it’s cool to work in IT at Tokić.