Arman Bedzhanian, BetConstruct AI: “We treat LLMs as a tool that our workers and managers can use to boost their productivity”

Shows & conferences - 13 November, 2026

Focus Gaming News was at SiGMA World 2026 in Rome, where it spoke with Arman Bedzhanian, Vice President at BetConstruct AI, about hyper-personalisation, the role of humans alongside AI and the practical use of AI in risk management.

Bedzhanian said personalisation in igaming predates the current AI hype, since AI is broader than LLMs and covers any model trained on data. The industry moved in the 2010s from rule-based engines to trained models, and is now moving from segment-based personalisation towards “hyper-personalisation”, in which a product and even a UI layout are tailored to an individual’s preferences rather than to a segment. The transition, he said, is not happening very fast: LLMs are adopted quickly for internal use, but using them in external products takes know-how.

On whether humans are still needed, he said replacing people is not BetConstruct’s target. The company has adopted many automations and uses LLMs to boost internal productivity, yet is still hiring and growing. “You do not need to replace people; you need to create tools for workers to make smarter and faster decisions,” he said, adding that AI is good at simple, routine tasks but not at strategic decisions or creativity.

In risk management, BetConstruct has shifted from human-driven to AI-driven pipelines, with models for detecting sports and casino fraud and bonus abuse. The next step, he explained, is an AI layer that filters out simple alerts and passes complex or sensitive cases, such as those involving high-value users, to risk managers, who can then focus on the hardest decisions.

 

His advice to small and mid-size operators was to start by looking at their data, including behavioural data, and to begin their AI transformation with internal operations, such as automating routine requests, withdrawals and deposit issues, which he believes would have the most impact on customer lifetime value.