DISTRIBUTED DATA ARCHITECTURES
Distributed data architectures have long been the holy grail of the ongoing data technology evolution, and it appears we’re getting close to that vision this year. An emerging approach to today’s data architecture focuses “not in what we do with data, but in where that data lives,” said Ari Weil, cloud evangelist and VP of marketing at Akamai Technologies. “We’re finally seeing a new generation of data platforms that live up to the vision that older models like data mesh and fog computing were reaching for. By spanning from the data center to the edge, they make it possible to store and use data where it’s actually needed—closer to users—without adding complexity.”
Minimizing data latency “has long been recognized as a key factor in improving user experience and increasing conversions,” Weil added, calling it “accelerating time-to-first-byte for dynamic content.”
Highly distributed data platforms “are also solving some of the toughest, longest-standing infrastructure headaches,” he added. For example, in inventory systems, “Businesses can now track and update product availability in real time, closer to where orders are placed, instead of relying on slower, centralized databases that often introduce delays or mismatches.”
Plus, on the operations side, “Teams are spotting and resolving outages faster because diagnostic data can be processed right where it’s generated, without waiting for it to move across the network,” Weil said.
AGENTIC AI
Within the past year, there’s been plenty of discussion and speculation about the potential of agentic AI, but how does it play out in the data space? “Unlike traditional machine learning models that respond only to predefined inputs, agentic systems are designed to operate autonomously,” said Don Woodlock, head of global healthcare solutions at InterSystems. This includes functions such as “setting goals, making decisions, and taking action with human review at a plan level instead of constant human prompting. It marks a fundamental shift from AI as a passive tool to AI as an active collaborator, capable of navigating complex, real-world environments with intent.”
Agentic AI has the potential “to shift AI from being reactive to genuinely proactive, anticipating needs before they arise and closing the loop between data and action,” Woodlock added.
Agentic AI has the potential to deliver significantly greater gains than incremental efficiency gains—it “marks a fundamental evolution in how businesses unlock value from the data they already possess and continue to generate,” agreed Bennie Grant, COO at Percona. “Agentic AI empowers organizations to move beyond passive data collection toward intelligent, autonomous decision making. Data becomes a living asset—constantly analyzed, interpreted, and acted upon in real time.”
Healthcare data sites could potentially see powerful benefits with the rollout of AI agents. Such systems “could continuously analyze patient data and take proactive steps in clinical settings such as hospitals or outpatient care,” said Woodlock. “For instance, they could recommend lab work, notify care teams, or prepare pre-surgical materials without waiting for explicit human input at each detail.”
In addition, “An agentic system might automatically coordinate tasks ahead of a scheduled procedure, ensuring required tests are completed and patient education and consent forms are delivered,” Woodlock added. “This level of automation can reduce clinician workload, accelerate response times, and support more consistent, scalable care. As these systems gain autonomy, they enable real-time management of complex workflows at a scale that is difficult to achieve manually.”
The caveats to moving deeply into agentic AI architectures are the need to boost greater transparency, safety, and accountability, said Woodlock. “These systems need to explain their decisions, build trust with users, and be governed in a way that ensures human oversight. There’s also the matter of regulation and testing, especially in high-stakes fields like healthcare.”
Additional challenges with agentic AI involve security, governance, and trust. Grant stated, “Companies need to make sure they’re staying compliant with evolving regulations and handling data responsibly.” While this is a standard concern with any technology, “The stakes are higher here because agentic systems are autonomous by design, meaning they make decisions and act with minimal human oversight. This raises critical questions about accountability, control, and risk management. Letting an AI just do things autonomously requires a level of confidence that most organizations aren’t ready to grant, at least not yet. It will take time—likely years—before trust is fully gained.”