In 3-5 Years, It Will Be Difficult for Operators and Platforms to Compete Without AI" — Oleg Smolerov

Note: The original version is available through this link in C-lvl telegram channel.

Oleg Smolerov is the CPO of ZingBrain AI , an innovative company transforming the iGaming industry through artificial intelligence. It improves key metrics, enhances player loyalty, and elevates the overall gaming experience.

Oleg shared insights on how personalization and player behavior prediction using machine learning help brands not only retain their users but also increase their GGR (Gross Gaming Revenue) by significant percentages.

C-lvl: How does the implementation of AI in iGaming companies affect traffic and profitability? On average, how much can a company’s GGR increase with AI?

Oleg: AI can be used in various ways, leading to different GGR growth outcomes. There are many options for integration, and each should be considered individually. For example, personalizing the lobby — our main product — can provide an additional 5-15% GGR. Players become more attached to the brand, sessions are longer and more frequent, and withdrawals decrease because they always find something interesting to play.

Using ML models, we can effectively predict churn rate, LTV, and identify bonus hunters early on. The GGR growth from these activities varies depending on how well the partner's processes were set up before implementing AI. While churn, LTV, and bonus hunters can be managed with simple algorithms, ML models still deliver more accurate predictions.

C-lvl: What is ZingBrain, and what does it do? How does it benefit users?

Oleg: Our main focus right now is lobby personalization. Every player sees games and sections tailored to their preferences, previous gaming experiences, and responses to past recommendations. Essentially, we do for casino lobbies what Netflix and Spotify do for their homepages.

With our models, we can present players with various game selections:

- Games with similar mechanics to their favorites

- Sorting new games by relevance to the player

- Games they haven’t tried yet but are likely to enjoy

- Games with their favorite mechanics, and more.

This ensures players always find something new and engaging, avoiding endless grids of games that are irrelevant to them. The lobby even adapts to the player almost in real-time. Once they experience this level of personalization, it’s hard for them to switch to casinos without it, increasing brand loyalty.

Moreover, we offer casino managers a Smart Push feature that allows promoting specific game sets through recommendations. This could include proprietary games or games from providers offering discounts for top lobby placement. We show these games to relevant players based on their preferences, continuously rotating them based on responses, which can significantly boost GGR (from 40% to 1000%) for selected games, providing extra profit for operators.

We also predict churn rates before players leave, allowing operators to react earlier and communicate with players before they churn. This increases the chances of retaining players, which is particularly important for high rollers.

Additionally, we predict LTV early on, which can be used to assess the effectiveness of traffic sources and identify high rollers early to avoid missing important players.

C-lvl: Why did you decide to work in the iGaming niche?

Oleg: We have extensive experience developing B2B products for iGaming. It's a large and rapidly growing industry, which we understand well and see many exciting opportunities for development. Around four years ago, we saw the potential of integrating AI and started developing the ZingBrain product.

C-lvl: Have you played gambling games yourself to better understand player preferences?

Oleg: Yes, I’ve been betting on sports since my youth and was a professional player in daily fantasy. For our product, we constantly research player behavioral patterns in casinos and test recommendations ourselves. Understanding player psychology and needs is essential for building a good recommendation system. For example, you can’t simply take recommendation models from e-commerce and apply them to casino games — it doesn’t work because gambling motivations are entirely different.

C-lvl: In your experience, is there a country that significantly lags in technology compared to competitors in iGaming?

Oleg: I don’t think it’s fair to attribute it to specific countries. We live in a globalized world, and everything depends on the people and teams involved. You can build a great product from any country if you have enough experience, knowledge, and motivation.

C-lvl: Some operators are still cautious about integrating AI into their platforms. How long do you think it will take for skeptics to change their minds and recognize the value of AI in iGaming?

Oleg: The situation today compared to two years ago is already a big difference. The global hype around AI is doing its part. But caution still exists, primarily due to two reasons: lack of understanding and unwillingness to share data. We actively help with the first issue by showing, through our analytics, the effects that can be achieved. We have plenty of documentation on how the product works, why it’s effective, and how it positively impacts both players and GGR. We've worked hard to simplify integration to take just a few weeks and be understandable to developers and product teams. Now, after studying all the information, partners rarely have questions.

Regarding data sharing, it's also manageable. We don’t use personal player data, and we don’t have direct access to the partner's UI/frontend to collect all player data. We only work with what the operator sends us. During integration, we do not receive personal player data and work only at the player ID level, supporting attribute anonymization through various techniques. Therefore, we may not even have explicit data on bet amounts, currencies, or geographies — everything can be anonymized and normalized without affecting the quality of recommendations.

In a year or two, AI will be present on all platforms, starting with tier-2. The benefits it offers are too significant to ignore. The simplest way to see its value is to try it and analyze the results. For example, conducting an A/B test. Usually, after analyzing the results, any doubts or caution disappear.

We offer something new, so we understand when partners are hesitant because they care about their products. Transparent communication is crucial here, which is why personal meetings help. We attend all key exhibitions and, being based in Cyprus with a large iGaming community, we frequently engage in discussions. We have several major partners whose founders are from the CIS, and we regularly communicate, jointly analyze results across different markets, and with each passing month, we notice a shift from skepticism to full trust.

C-lvl: Are there any differences in working with clients from CIS markets and Western markets? Which regions are more open to AI?

Oleg: As I mentioned, I believe in people, not nationalities. Over the past 18 months, I’ve done over 200 demos, and it's usually clear during the first call whether we’ll work together based on the depth of questions and reactions to answers. There isn’t a radical difference in understanding AI across regions.

With CIS partners, discussions are often easier due to similar mentalities, and there is a noticeable focus on results and understanding what can be achieved with AI that cannot be done otherwise. The British tend to follow a more bureaucratic approach and gather more information before making a decision. Still, if they understand the benefits of AI, they’re willing to invest time in navigating all bureaucratic steps. I encountered slightly less understanding in LATAM, but that’s expected, as the market is young, and their focus is different. It’s also crucial there to be introduced by someone local whom they know and trust. However, these are generalizations.

Any specific person or team can make decisions quickly if they know what they want, are satisfied with what they hear during the demo, and are on the same wavelength as us. I’ve heard a lot about the specifics of the Asian market, particularly Southeast Asia, but we have a partner there who integrated faster than anyone else, constantly makes improvements, and is open to any communication.

C-lvl: How can you reassure players concerned about AI stealing their data for fraudulent schemes?

Oleg: How do Netflix, Spotify, YouTube, etc., explain it to their users? All tech giants actively use AI to understand preferences, offer relevant content/ads/products. We’re much more harmless in this regard. We don’t use personal player data, nor do we have direct access to the partner's UI. Essentially, we only know that Player A made bets in certain games — at most, their country and language. This information is not sufficient for fraudulent schemes but enough to improve their gaming experience.

C-lvl: Can AI replace any employees in the iGaming industry? If so, which ones?

Oleg: I wouldn’t say replacement. My favorite example is typewriters, printers, scanners, etc. Did they eliminate jobs? No, they just made it possible to complete much more work in a shorter time, and the amount of work increased significantly. The same applies to automobiles — people simply started moving more because they had more opportunities, and the number of jobs actually grew.

For our product, we don’t replace casino managers. We provide them with a tool that allows them to work faster and better, with less routine manual labor and better results. It frees them up for more interesting tasks: deeper player analysis, new features, market research, promotional campaigns, closer communication with providers, etc.

In player analytics, it’s also about improvement rather than replacement. ML predictions for churn and LTV are more accurate, enabling more effective player segmentation and better targeting in bonus campaigns.

However, I do expect staff reductions in support departments, which

are already happening. Implementing good AI support and chatbots is much cheaper and more efficient than maintaining dozens or hundreds of support staff. Only supervision and time for model training are needed.

C-lvl: What are the future prospects for AI in iGaming?

Oleg: The potential is huge. Practically every step of the player journey in a casino or sportsbook can be optimized with AI. In 3-5 years, it will be very challenging for operators and platforms without AI to compete.

AI and data-driven approaches demonstrably outperform any manual settings or algorithms. Imagine a brand where:

- Welcome bonuses are automatically tailored for cohorts based on traffic sources and player geolocation

- Traffic from each source is instantly scored for effectiveness

- Players see only relevant content from their first interaction, which adapts in real time based on behavior

- All promotional campaigns run automatically, supervised by managers according to budgets

- Each player has a predicted LTV and churn probability. If there is a high risk of churn, they are re-engaged through the most effective channel with a suitable bonus

Now compare this with a brand where everything, including analytics, is done manually. The ROI is incomparable, and products that don’t keep up with the times will slowly but surely fade away.

C-lvl: Which events on AI development would you recommend for anyone interested in the niche?

Oleg: AI is just starting to penetrate iGaming. There are almost no specialized events. There are meetups and podcasts touching on the topic, but I wouldn’t recommend anything specific. It’s better to visit our booth — we attend all major exhibitions, and we’d be happy to chat and share our experiences.

Thank you for your attention!