The technology industry has pushed cloud-first narrative too much in enterprise communications and contact centers, Avaya’s CEO claims.
The on-prem vs cloud debate in CX technology has been frequently debated on in the past few years, with many notable vendors having promoted cloud as the long-term solution, while many point out that organizations continue to rely on on-prem infrastructure or hybrid models.
The discussion reflects a wider industry shift as enterprises balance innovation in cloud platforms with existing infrastructure, regulatory requirements, and long-term technology investments.
On LinkedIn, Patrick Dennis, CEO of Avaya, claims that the industry has oversold cloud-only CX platforms, and that the strategy is failing in the AI era.
“For years, the industry narrative has been shoved down our throats: Move every contact center seat to the public cloud or get left behind,” he explained.
“While the world rushed toward multi-tenant CCaaS, a massive number of enterprise seats stayed put. Why? Because performance guardrails, latency, and proprietary data privacy aren't "legacy concerns"—they’re competitive moats.”
Cloud-Only Contact Centers Face Data and Performance Limits
Dennis argues that the technology industry is expected to see a “hybrid resurgence” by combining both cloud services with on-prem infrastructure rather than relying entirely on multi-tenant cloud platforms.
This stems from several practical limitations in cloud-only architectures, with data and property risks becoming more significant in the AI era, feeding enterprise data into public cloud models can expose sensitive information for companies that depend on proprietary data.
“There’s literally nothing you can do about that metadata leakage. For a serious enterprise, that’s a non-starter,” he wrote.
He further highlights the latency and performance issues in cloud platforms, particularly with contact centers that use voiced-based AI systems in their operations.
Due to the high-performance nature of AI voice systems, they demand near-instantaneous processing, and voice traffic routed through multiple public cloud services can create severe delays, making private or on-prem infrastructure necessary to keep them close to the data.
“Voice AI demands near-zero latency. If you’re routing a voice stream through three different public cloud hops, the conversation falls apart,” Dennis continued.
“High-performance AI requires getting the compute as close to the data as possible. You can't cheat physics with a SaaS subscription.”
This argument reveals that the industry will likely see a combined cloud service with on-prem infrastructure, arguing that analysts who predict a fully cloud-based future are overlooking these technical and security restraints.
In fact, new demands from AI, voice processing, and enterprise data security may strengthen the popularity of these hybrid architectures.
Why the Technology Industry Promoted Cloud-First CX Strategies
Having promoted cloud adoption in enterprise communications and contact centers for more than a decade, the technology industry has solved several limitations of traditional on-prem systems and created new opportunities for vendors.
These included lower upfront costs, faster deployment, and easier scalability compared with traditional on-prem infrastructure, as enterprises began switching to platforms such as CCaaS and UCaaS to avoid large hardware investments and continuous software updates.
For features like analytics, automation, and AI, the model enabled faster innovation and supported remote agents and distributed workforces during the COVID-19 pandemic, which on-prem systems found difficult to do.
As a result, many vendors and analysts have begun promoting cloud-first strategies as a long-term direction for enterprise communications.