Delasport CEO Oren Cohen Shwartz explains how AI and automation are reshaping the company’s operating model and changing the economics of growth.
Delasport is a leading iGaming supplier known for product innovation and challenging industry conventions. In recent years, the company has introduced products such as Betiator and My Sportsbook personalization while expanding into more than ten regulated markets, including the UK, the Netherlands and Ontario.
Gaming Intelligence recently sat down with Oren Cohen Shwartz, Delasport’s CEO for the past five years, to discuss another transformation taking place behind the scenes: how AI and automation are reshaping the company’s operating model and changing the economics of growth.
Delasport has been an early AI adopter. What triggered it?
Growth, actually. Delasport expanded very quickly. More partners, products and markets naturally required more people and processes to support them. But rapid growth creates complexity and traditionally requires additional resources. You build structures for a particular stage of the company, expecting the business to continue growing and scaling.
Then came AI. Around two years ago, we began systematically examining how it could be integrated across the organization. We asked ourselves: What can we simplify or automate? Where can we reduce manual work? How can we increase productivity and quality while continuing to expand?
How did you begin the transformation?
We started by bringing in an AI consulting company specializing in the organizational adoption of AI. We wanted a quick start, and the engagement allowed us to identify areas where we could achieve rapid improvements using existing AI tools and techniques. They interviewed stakeholders from multiple departments, including R&D, Product, Marketing, Customer Support and CRM, to identify patterns, bottlenecks and repetitive manual work.
The objective was not simply to do the same work faster. It was to improve what we build, make the organization more intelligent and scale more efficiently. The consultants accelerated the process because they knew how to identify the activities where AI could create the greatest value and which existing off-the-shelf tools could address them.
That was the starting point. From there, the challenge was to develop the internal knowledge and ownership required to turn individual use cases into sustainable organizational capabilities.
What was the goal of this transformation?
The goal of Delasport’s AI transformation is to increase the speed, quality and intelligence of everything we build and operate, allowing us to deliver greater value to operators and players while scaling the business more efficiently.
For us, AI is not a standalone initiative or a feature added to make an existing product sound more innovative. It is becoming part of the infrastructure through which we build products, operate the business and support our partners.
From your experience what pitfalls can companies face during an AI transformation?
There are several types of pitfalls. One of the most common is treating AI adoption as simply giving employees access to tools and expecting immediate productivity gains. Without proper governance, training and measurement, companies can increase costs and risks while producing more output but of lower quality.
In R&D, for example, AI coding tools are extremely powerful, but they need to be managed. Different models have different capabilities and costs. Using an expensive, advanced model for a simple task wastes resources, while using a weaker model for a complex task can generate poor code and create additional rework. Weak prompts or insufficient context can also send developers into a loop of repeatedly correcting bad output, consuming tokens without making real progress.
AI-generated code must also meet the same architecture, security, testing and review standards as code written without AI. Otherwise, companies may generate technical debt and vulnerabilities faster than they generate genuine value.
Another mistake is measuring success by AI usage rather than business outcomes. The relevant questions are not how many employees use AI or how many lines of code it generates, but whether development cycles are shorter, quality has improved, incidents have decreased and customers are receiving value faster.
Another pitfall is choosing the wrong AI tool. Some off-the-shelf tools do not meet the organization’s needs, while others may work well but become prohibitively expensive at scale. In some cases, it is better to develop the capability in-house, provided the R&D investment makes commercial sense. This can provide a solution tailored to the company’s specific needs while reducing dependency on recurring third-party fees.
Other mistakes include expecting employees to use AI effectively without training them in prompting, context management and output validation, or creating many disconnected AI experiments that never become scalable, production-grade capabilities.
Companies must also protect their source code, intellectual property, customer information and regulated player data. Employees need clear rules defining which tools are approved and what information may be shared with them.
In short, AI tools need to be governed and “tamed.” Companies should not merely push employees to use them; they must teach people how to choose the right model, provide the right context, validate the output, protect sensitive information and measure the actual business value created.
Which leads to the question everyone is asking: is AI replacing people?
AI is absolutely changing jobs, but reducing the discussion to “AI replaces people” misses what is really happening.
Repetitive, rules-based activities can increasingly be automated. In many cases, however, the role does not disappear, its composition changes. Analysts spend less time assembling information and more time interpreting it. Engineers can focus more on architecture and complex problem-solving. Product teams can validate ideas faster. Judgement, creativity, specialist knowledge and critical thinking become more valuable.
But we should also be realistic: when technology changes how work gets done, organizational structures and resource requirements change with it. AI coding tools enable companies to develop products faster, while AI-driven automation can perform parts of testing more quickly and at a greater scale than traditional manual QA. This may mean that fewer people are needed for certain activities, while demand increases for people with different or more advanced capabilities.
The more accurate conclusion is that AI is replacing some tasks, reshaping many roles and creating new ones. Increasingly, people who know how to work effectively with AI will have an advantage over those who do not.
Has this resulted in organizational changes at Delasport?
Yes. Some roles and responsibilities have changed, and some positions have become redundant. Those are difficult decisions because there are people behind every one of them, and we do not dismiss that.
At Delasport, the objective is not simply to reduce headcount. It is to increase the capability of each team, improve quality and accelerate delivery. AI should allow us to grow and produce significantly more without increasing headcount at the same rate.
For a long time, adding headcount was almost treated as evidence of growth in technology companies. I think that equation is becoming outdated. The more interesting question today is how much additional business an organization can absorb without adding resources at the same rate.
In the AI era, operational capacity per employee may become a much more meaningful growth metric than headcount itself. Our objective is to handle significantly more business without allowing organizational complexity and cost to grow at the same rate.
What does a more efficient Delasport mean for your partners?
It means faster delivery, better-performing products and more scalable operations. We are introducing improvements across our core Sportsbook, Casino and PAM platforms – from higher conversion and smarter player journeys to stronger trading and risk capabilities and the automation of repetitive operational work.
It should also help us launch partners and enter new markets faster, respond to their requirements more effectively and continuously improve the performance of their businesses. Ultimately, our internal efficiency must translate into measurable value for our partners and their players.
What do Delasport for the next few years look like?
Over the past couple of years, we have launched dozens of brands across more than ten regulated markets, including the UK, the Netherlands, Sweden, Denmark, Brazil, the Philippines and Ontario, and we are now pursuing opportunities in markets such as Alberta.
We have won numerous industry awards for product excellence and introduced genuinely innovative products into an industry that has seen relatively little fundamental change for close to a decade. Behind those headline innovations is continuous development designed to make our core products perform better.
Our ambition for the next few years is therefore clear: to continue innovating, become significantly more efficient and scale the business without allowing cost and complexity to grow at the same pace.
Partner Editorial