Here’s a concrete data point for the “agentic AI” hype cycle. Floward, the online flowers and gifting company that operates across the Middle East and UK, says it now handles up to 13x higher peak-day conversation volumes after rebuilding its customer service on Infobip’s AgentOS platform — moving from rule-based chatbots to a multi-agent AI system.
The problem Floward was solving is brutal and specific. As a same-day delivery business built around occasions — Mother’s Day, Valentine’s Day, Ramadan — it faces demand spikes that would swamp a human-only support team. Working with Infobip’s consultants, it deployed a system that routes conversations to specialized AI agents for tasks like address collection, FAQs and order changes, escalating to live agents only when needed. A neat touch: orchestrated WhatsApp journeys that walk gift recipients through address collection in chat. The whole thing was built and deployed in under two months.
The numbers Floward is putting forward
Floward’s claimed results are the headline: a 15% cut in customer-service costs, 54,000 conversations handled on Valentine’s Day alone, one-minute response times at 95% SLA, and a 12-percentage-point jump in customer satisfaction — during its busiest periods, not its quietest.
“One of the biggest achievements of this transformation was proving that scaling customer service no longer means scaling headcount at the same rate.”
Lujain Mallosh, Customer Care Senior Manager at Floward
These are, of course, vendor-supplied figures from a case study Infobip is actively promoting, and “AI containment” metrics deserve scrutiny. But the underlying story — absorbing a holiday-scale surge without a matching surge in staff — is exactly the enterprise promise that has made agentic customer service one of 2026’s most-hyped AI categories. Emir Kalem, Infobip’s Head of Customer Success for EMEA, called Floward’s peak-season crunch “a textbook example of where agentic AI delivers the most value.”
