Maksym Shtun from Slotegrator discusses AI: strengths, limits, and future possibilities

In a discussion with iGamingExpert, Slotegrator Product Owner, Maksym Shtun, examines the current status of AI adoption within the iGaming industry, highlighting the risks of improper use and detailing the company’s strategies for integrating AI into its operations.

How would you describe the overall usage of AI within the industry at the moment? Is its full potential being realised?

The industry’s approach to AI is currently transitioning from observation to active integration. The sector has certainly become more daring and proactive in experimenting with these technologies compared to previous years.

However, we are still a long way from realizing its full potential. Currently, there is a significant disparity between technical capability and practical application.

This is primarily due to ongoing concerns regarding data security, the necessity for robust human oversight, and a deeper comprehension of the technology’s long-term consequences. The industry is advancing rapidly, but it is doing so with a necessary degree of caution to ensure stability and trust.

Is there a danger of leaning into AI too much in a way that could potentially be detrimental to the overall product?

The danger lies not in using AI, but in misunderstanding its nature. Some individuals perceive it as a deterministic system, akin to a calculator that consistently yields the same result for identical inputs. However, AI does not operate in this manner. There is inherent variability in its responses, and this deviation intensifies when decisions are made sequentially.

The moment a team delegates 100% of critical decisions to AI—whether it involves content generation, logic handling, or user-facing outputs—they forfeit the ability to guarantee a consistent experience. And consistency is what users rely on.

The appropriate approach is to regard AI as a capable collaborator, not an autonomous operator. Humans must still establish the parameters, verify the outputs, and assume responsibility for the outcomes. If an issue arises and you cannot explain the AI’s decision-making process, you have already lost control of your product. This is not an AI issue, but a process issue.

How does Slotegrator currently utilise this technology, and do you plan to step this up? If so, in what way?

Our current methodology is highly pragmatic. We recognize significant value in employing LLMs in conjunction with RAG (Retrieval-Augmented Generation) architectures. This enables us to implement effective, streamlined solutions for processing complex documentation and transforming static knowledge bases into interactive, highly accurate resources.

In addition to data processing, we are integrating AI into our creative workflows. It has become an indispensable tool for rapid prototyping, generating mockups, and exploring visual concepts, thereby significantly accelerating the ‘ideation-to-execution’ cycle without supplanting human oversight.

Looking ahead, we acknowledge the potential of MCP (Model Context Protocol) solutions for fostering more interconnected environments. However, our interest in this area is highly specific: we view it as a means to enable tightly defined, isolated functionalities for particular tasks.

For us, the focus is on maintaining complete transparency and ensuring that any AI integration operates within a strictly defined perimeter. It is not about broad accessibility, but about precision and control over how the technology interacts with our internal tools.

What role does client feedback play in shaping how you utilise not just new tech like AI, but your overall roadmap?

Client feedback serves as a crucial indicator for potential optimization. When we observe recurring inquiries or similar requests for specific functionalities, it suggests that there may be a broader opportunity to enhance the user experience.

However, we approach this with an emphasis on suitability. First, we assess whether a process genuinely benefits from automation and if it aligns with how our partners interact with the system.

If automation is the appropriate course of action, we then determine the most effective tool for the task. AI is a potent option for managing complex or unstructured data, but we only apply it where it offers a distinct advantage over traditional methods.

Our roadmap is shaped by this pragmatic approach: identifying the correct solution for each specific scenario, whether that involves AI or a more conventional technical enhancement.

Following on from the previous question, what else can we expect from Slotegrator as the company keeps driving forward?

What you can anticipate from us is a continuous effort to provide operators with superior tools for making more informed, quicker decisions. This is the underlying theme of all our current development efforts.

With Casino Builder, the objective is to shorten the timeframe between conceptualization and implementation. We are incorporating AI-driven features that assist operators in configuring, adjusting, and refining their platforms with significantly greater precision—enabling decisions that previously required days of back-and-forth to be made confidently in a fraction of the time.

Beyond the builder itself, we are making substantial investments in how operators access information and guidance. Static documentation has its limitations. We aim to replace it with intelligent, context-aware support that delivers the correct answer at the opportune moment—so operators are not guessing, but acting on accurate, timely insights.

The overarching direction is as follows: every feature we release should either reduce friction in an operator’s workflow or enhance the quality of the decisions they make. Speed is important, but speed without accuracy leads to downstream problems.

Our objective is to achieve both: empower operators with the confidence to act swiftly because the tools they utilize are dependable and designed around their actual operational methods.