We are moving away from cohort-level personalisation to a player-specific experience, real-time, on-device, and privacy-conscious, said Navneeth Srinivas, Founder of Metamine Gaming, during one of the most anticipated panels at the SiGMA Asia 2025 summit in Manila.
In a session titled AIs High-Stakes Makeover: Utilising AI for Personalised 바카라er Engagement, industry leaders took a deep dive into how artificial intelligence is transforming customer engagement, player retention, and operational efficiency in the gaming sector.
Moderated by Tejinder Kumar, Group CCO at Kyrrex, the panel also featured Jack Wheeler, APAC Lead at Optimove, and Navneeth Srinivas, Founder of Metamine Gaming. Together, they unpacked the shift from generic segmentation to hyper-personalised player journeys.
“AI has done a decent job lumping us all into neat little boxes,” Kumar said, setting the tone for the discussion. But when do we move past male, 35, likes clicking stuff and actually start engaging players as unique individuals?
Wheeler highlighted how modern AI models can now predict user behaviour early in the customer lifecycle. You can understand if a player is going to be a low or high value player within the first 48 to 72 hours via AI, he explained. This early insight enables operators to customise bonuses, promotions, and content accordingly.
Srinivas introduced a compelling approach using Edge AI: We process everything on the device and not in the cloud… the players personal information does not leave the device. This approach supports compliance with the GDPR and Indias DPDP, ensuring real-time personalisation without compromising privacy.
He cited a fraud prevention use case in which edge AI identified potentially fraudulent behaviour and dynamically altered the users experience to prevent abuse, all without requiring human intervention.
Wheeler stressed that clean data is critical: Your data is only as good as where its coming from. He noted that many operators underestimate the need for structured, QA-tested data pipelines and the patience required to warm up an AI model, typically two to three months.
When measuring success, both experts emphasised the importance of A/B testing and tracking individual-level metrics, including revenue uplift, engagement time, and churn prediction. You’re basically looking back at two to four weeks of data to compare predicted results with actuals, said Wheeler.
The panel didnt shy away from the ethical side of personalisation. One of the biggest fears is over-personalisation that fuels compulsive behaviour, Srinivas warned. He advocated for including behavioural scientists in AI development teams to ensure models promote responsible gaming.
Wheeler cited an example with Stake.com, where AI identified high-risk, high-reward players likely to shift from sports betting to 바카라 games. Personalised, cross-channel outreach led to a 60C70% engagement rate, demonstrating the potential of targeted AI strategies.
Srinivas shared a campaign for a poker and rummy operator in India. We looked at where players spend time in-app, then offered personalised bonuses and even travel incentives, like one-way tickets to Bali. The result? A measurable uplift in player deposits and retention.
As AI models evolve, so does their ability to adapt across different regions and game types. Wheeler wrapped up by emphasising model adaptability: An AI model that is good should be able to identify behaviours and give individualised scores, whether a player logs in daily or plays poker once a month.
This panel highlighted the rapidly evolving intersection of AI, personalisation, and player responsibility. For anyone invested in the future of gaming, staying up-to-date with these developments isnt optional; its essential.
Dont miss out on the next big conversation. Follow the SiGMA Asia 2025 Agenda to stay informed on upcoming panels and speakers.