Building the digital backbone of healthcare: Cloud, resilience and AI at scale
By Manoj Warrier, VP, Infrastructure, and Sudhindranath Byna, Executive Director, Cybersecurity Architecture, Providence India
The technology that matters most in healthcare is usually the technology nobody notices.
A clinician opens a patient record, a scanner sends an image, an AI model supports a decision, and the experience feels effortless. Behind it is a connected chain of cloud platforms, networks, identity services, applications, data and security controls. Its complexity becomes visible only when a link breaks, often revealing how far the impact can travel.
As care becomes more digital, infrastructure has moved from a back-office function to the operating layer of the health system. It must scale, anticipate risk, and keep essential services running when something goes wrong. Three disciplines make that possible: cloud, resilience, and AI, with cybersecurity woven through all three.
From cloud migration to cloud maturity
Healthcare organizations have been migrating workloads for years, but the work does not end when an application moves to a cloud platform. Clinical and OEM systems still operate across different environments, with dependencies that may only become visible when something changes.
The scale of this challenge becomes clear when healthcare organizations undertake large-scale cloud transformation. In our experience, we observed this at scale across 51 hospitals and 1000+ clinics, with more than 32,000 servers decommissioned, 2,000+ applications retired and 650+ applications migrated to the cloud. The work involved understanding the systems around those applications, the identity services and devices they relied on, and what could happen to a clinical service if one dependency became unavailable. The broader lesson is that cloud maturity depends on how well teams understand and operate the environment they have built.
As healthcare environments become more distributed, technology teams need visibility into how services operate across cloud and on-premises environments, how dependencies interact and how changes can be introduced without creating disruption elsewhere. Cloud becomes more valuable when it gives teams the flexibility and resilience to operate that complexity with confidence.
AI moves into the operating layer
This becomes even more important as AI moves into the operating layer of technology. The conversation around healthcare AI often focuses on clinical applications, but infrastructure teams are finding practical opportunities as well. AI and automation can help make sense of the volume of logs, alerts and performance signals generated across modern environments, identify patterns earlier, automate routine work and support predictive maintenance.
That creates an opportunity to move technology resilience toward anticipation. AI can help teams find signals that might otherwise be difficult to identify across a large, distributed environment. But the technology still needs people around it. An unusual pattern could indicate a technical fault, a security incident or a legitimate change in usage. Understanding the difference requires context and experience.
The most useful model is one where AI assists and augments engineering teams while people retain responsibility for decisions that require judgment.
The operational value can be tangible. With automation, we observed approximately 80,000 hours of productivity savings. The significance goes beyond time saved. Reducing repetitive work and manual handoffs gives engineers more capacity to investigate complex issues and focus on work where experience matters most.
Building a resilient digital backbone
The same thinking applies to cybersecurity and business continuity. In healthcare, a security response cannot be considered separately from the clinical service it protects. Taking a compromised system offline may be appropriate, but the decision also needs to account for what happens to the people who depend on it while the issue is being resolved.
Recovery planning therefore needs to begin with the service a clinician or care team is trying to deliver. A diagnostic workflow may depend on an application, network connectivity, identity services, data feeds and an external platform. Restoring one component does not necessarily restore the service if another dependency remains unavailable. Testing realistic disruption scenarios can expose unclear ownership and dependencies that are difficult to see during normal operations.
This also changes how technology teams think about failure. Resilience is rarely about preventing every incident; complex environments will always have points of failure. It is about knowing which services matter most, understanding their dependencies, rehearsing how disruption will be managed and creating fallback paths that allow essential work to continue when systems are disrupted.
Healthcare will continue to add connected systems, digital services and AI-supported workflows. The technology environment will become more capable, and more complex.
For technology leaders, the priority is to make that complexity manageable. Cloud, cybersecurity, resilience and AI increasingly need to be considered together because each affects how the others perform. A strong digital backbone gives teams the visibility to understand dependencies, the resilience to contain disruption and the intelligence to identify problems earlier.
The goal is ultimately simple: give healthcare the capacity to evolve technologically while essential services continue to operate. When the foundation is designed with that continuity in mind, new capabilities can be introduced with greater confidence, while clinicians get the reliability they need to focus on care.