From scripts to Agentic AI: how Bally’s is redefining iGaming customer support

In this podcast, Zendesk’s Sarah Al-Hussaini (VP, AI GTM) sits down with Matt Litherland, Director of Service Delivery at Bally’s Interactive, for a candid look at how Bally’s evolved from scripted bots to Agentic AI—and what it really takes to make that shift successful.

In just six months, Bally’s tested and refined multiple AI models, achieving 25–30% deflection across messaging (which handles 90% of their contacts) while keeping human agents front and center for responsible gambling and complex cases. Matt shares how his team built trust in AI through baby steps, graceful failures, and co-designing with agents—turning change management into collaboration.

The results speak for themselves: customer contacts dropped from 7–8% to under 5% each month, agents now focus on higher-value interactions, and knowledge has transformed from static FAQs into conversational, AI-powered answers that resolve issues faster and strengthen player loyalty.

Key takeaways:

• Start small: begin with high-confidence use cases (like bonus requests or withdrawal status) to demonstrate quick wins.
• Empower your team: involve agents in co-designing flows to boost accuracy, trust, and adoption.
• Deflection ≠ deletion: let AI handle context so agents can resolve escalations faster and with more insight.
• Messaging leads the way: once optimized, expand success to channels like email auto-replies and agent assist.

…and many more. To find out, listen now.