Banks have made significant progress with cloud, but operational complexity, regulatory pressure and decision bottlenecks are limiting speed and impact.
In this exclusive interview with The Digital Banker, Salma Datenis, VP and Head of Cloud Studio at Amdocs, explains why financial institutions are already rethinking their cloud operating models and how Agentic AI can be part of the solution.
The Digital Banker: Cloud adoption in financial services has accelerated over the past decade. Where are banks today and how is the broader context around cloud and AI evolving?
Salma Datenis, Amdocs: Most banks today are well beyond the question of whether to adopt cloud. Cloud is now foundational in supporting modernisation, digital product delivery, customer experience, cost discipline, and operational resilience.
What has changed is the scale and ambition of what banks now expect from cloud. As cloud estates have grown larger and more interconnected, traditional operating models are being stretched. Automation has helped, but it is no longer sufficient to deliver the speed, efficiency, and control that banks need.
In that context, AI is increasingly being leveraged to strengthen how cloud environments are run. Rather than being used for standalone single use cases, AI is now being embedded into the core of cloud platforms and operations helping banks manage complexity, improve economics, and support faster, more confident change. Boards no longer see cloud as just a place to run workloads, but as a strategic capability that must operate intelligently at scale.
What challenges are banks facing as they scale cloud – particularly in cloud operations?
The core issue is that cloud complexity is growing faster than operating models. Multi-cloud fragmentation, wasted spend, tool sprawl, and manual effort are all increasing even in environments that appear highly automated. Traditional automation helped banks get started, but it struggles as estates become larger, more distributed, and more dynamic.
Many cloud operations teams are now hitting an automation plateau. As automation scales, so do exceptions, dependencies, and coordination overhead. The result is diminishing returns: more scripts and tools, but slower execution and higher cognitive load.
This is fundamentally a cloud-operations challenge. Modern cloud environments require continuous decisions around cost, performance, risk, and change and those decisions still rely heavily on human coordination. That is where speed, consistency, and confidence begin to break down.
So, what needs to change in how banks run cloud and where does Agentic AI come in?
What needs to be changed is the cloud operating model, not just the tooling. In the agentic AI era, banks have the opportunity to move beyond task automation towards AI-driven execution that handles complexity more effectively.
Agentic AI introduces autonomous, outcome-driven capabilities that improve responsiveness and scalability in cloud operations particularly where traditional automation struggles with exceptions and judgment-heavy decisions.
At Amdocs, we refer to this shift as Agentic Cloud: a cloud operating model where AI agents coordinate decisions and actions across cloud environments, within clearly defined governance boundaries. The objective is not unchecked autonomy, but faster and more consistent execution with auditability, and control embedded by design.
Where do banks see the most value applying Agentic AI in cloud operations today?
Banks that are making progress are highly selective. They focus on high-impact cloud-operations workflows where manual coordination slows execution and increases risk.
Typical starting points include:
• Day-to-day cloud operations, where teams struggle to move from monitoring and alerts to timely, and even proactive, action
• Cloud cost and FinOps optimization, where inefficiencies are visible but hard to address continuously
• Governance-heavy cloud changes, where every action must be explainable, auditable, and reversible
• Cloud migration and modernization operations, where discovery, dependency mapping, and sequencing slow delivery
Research shows banks are cautious early on, prioritising lower-risk use cases. This selective approach allows institutions to build confidence, experience and governance before expanding into higher-impact scenarios.
If banks modernise cloud operations successfully, what impact does that have beyond IT?
The impact is enterprise-wide, because cloud operations underpin how quickly and safely a bank can change. In a recent Amdocs research with Coleman Parkes, 70% of banks said delaying AI-driven cloud operations would harm their competitive position, reinforcing that this is no longer viewed as a purely technical issue.
Improved cloud execution directly affects cost control, resilience, compliance, and customer experience areas that matter to every CXO. It also reduces operational friction, allowing teams to spend less time coordinating and more time delivering outcomes.
The differentiator will be governance. While most banks are formalising standards and Centers of Excellence, far fewer have them fully operational. Institutions that align cloud, data, and governance effectively will be best positioned to scale Agentic AI safely and turn cloud operations into a sustained advantage.
