As a specialist in Business Intelligence and data science, Chloe Maraina has spent her career translating complex datasets into visual narratives that drive organizational change. Her expertise lies at the intersection of data management and strategic integration, where she currently observes how the next generation of artificial intelligence is reshaping the very foundations of the IT department. With a deep understanding of how infrastructure and automation converge, Maraina offers a unique perspective on the shift from manual scripting to the era of autonomous, agentic workflows that are set to redefine the digital landscape by 2026.
The following discussion explores the rapid evolution of automation, moving beyond basic task execution to a future defined by self-optimizing systems and agentic AI. The conversation delves into the surge of investments in cloud and infrastructure automation, the emerging dominance of AI-assisted code which is expected to comprise the majority of development by 2027, and the democratization of technical skills through natural language “vibe coding.” Central to this transformation is the rise of multi-tiered AI agents—ranging from task-specific solvers to orchestrating executive agents—and the persistent challenge of overcoming siloed data to achieve a truly autonomous enterprise.
With the cost of automation decreasing, which niche IT tasks are now the most viable for transformation?
The financial barrier to entry for high-level automation has crumbled, allowing us to look at the “long tail” of IT operations that were previously ignored because the ROI just wasn’t there. We are moving into a phase where nearly 64% of IT professionals are aggressively investing in cloud automation, and 50% are prioritizing workload automation and service orchestration. In the past, small-scale configuration tasks or specific infrastructure tweaks were handled manually because writing a custom script took more time than the task itself. Now, because tools have become faster and more intuitive, we can justify automating these micro-processes, which collectively removes the “death by a thousand cuts” that many IT teams feel. You can see this reflected in the 49% of organizations now pouring resources into infrastructure and DevOps automation, focusing on those granular, repetitive actions that keep the lights on but drain human energy.
Beyond mere speed, how is the shift toward autonomous decision-making fundamentally changing the day-to-day operations of IT departments?
We are witnessing a pivot where AI is no longer just a digital hands-on-deck but is actually taking over the steering wheel for complex decisions. IT teams are now deploying algorithms that don’t just alert a human to a problem but independently reallocate network resources to prevent a bottleneck before it even manifests. Imagine a system that monitors traffic patterns in real-time and makes provisioning and deployment decisions without a single ticket being opened; it removes the “human-in-the-loop” lag that often leads to downtime. This shift toward autonomous configuration and deployment is significantly boosting accuracy and consistency, as the AI doesn’t get fatigued or overlook a minor setting during a midnight server migration. It creates a sensory environment where the infrastructure feels almost “alive,” constantly breathing and adjusting itself to meet the demands of the business.
We are seeing a significant move toward “agentic AI.” How are these agents evolving from simple task-handlers into multi-stage, cross-functional collaborators?
The evolution toward agentic AI is perhaps the most exciting trend we’ve tracked, with 57% of organizations already using agents to handle multistage workflows. We are moving away from “if-this-then-that” logic toward systems that can detect an anomaly, diagnose the root cause, and execute a multi-step remediation plan autonomously. While only about 16% of companies have reached the point of end-to-end processes that span multiple teams, the ambition is clearly there, with 81% of organizations looking beyond simple task automation. These agents are categorized into distinct tiers: solver agents for specific tasks, worker agents that collaborate with humans, and the upcoming executive agents that will manage the entire ecosystem. The ultimate goal is a system where you provide a high-level objective, and the AI agents invent the necessary tools and processes to achieve it, acting with a level of authority that was once reserved for senior IT leadership.
The software development life cycle is undergoing a radical transition. What does the data tell us about the future of code generation and the role of the traditional developer?
The era of “handcrafted” systems is rapidly fading as we move toward agentic, goal-driven applications that translate human intent directly into code. Current data shows that AI-assisted code already accounts for 42% of all committed code, and that number is projected to climb to a staggering 65% by 2027. This doesn’t mean developers are becoming obsolete, but their role is shifting from syntax experts to architects of intent, where they supervise AI agents that write, test, and secure code autonomously. Software development has seen the biggest impact from agentic AI, with 57% of organizations reporting significant improvements in this functional area. The feeling in the dev shop is changing from one of manual labor to one of high-level orchestration, where the pipeline is a living entity that constantly refines itself to be more secure and efficient.
“Vibe coding” has become a buzzword recently. How is this “plain language” approach to development expanding the pool of creators within a company?
Vibe coding is the ultimate expression of democratized technology, where the barrier between a business idea and a working application is reduced to a simple conversation. Instead of needing to understand API connections or complex architecture, a user can describe their desired outcome in plain English, and the AI handles the technical execution. This allows people across customer service, which has seen a 55% impact from AI, and marketing, at 46%, to automate their own daily responsibilities without waiting for a developer’s help. It changes the atmosphere of the workplace from one of technical silos to one of universal creativity, where a supply chain manager can “vibe” an automation tool into existence to solve a logistics bottleneck. It empowers every employee to be an active participant in the company’s digital transformation, effectively turning the entire workforce into a decentralized IT department.
Despite these technological leaps, data remains a massive hurdle. What specific obstacles are preventing organizations from reaching the “executive agent” tier of automation?
The reality is that even the most advanced AI agent is only as good as the data it can access, and right now, data is often scattered and siloed across the enterprise. About 42% of professionals admit that they face significant data quality and access issues when trying to deploy these intelligent agents. Without a unified “control plane” and robust governance, we cannot yet move to the “executive agent” level where AI manages other AI. Many organizations have spent years building fragmented systems that don’t talk to each other, creating a sensory landscape of “data graveyards” that agents simply cannot navigate. To reach that next tier of self-optimizing infrastructure, we have to do the heavy lifting of cleaning and integrating our data sets so that the AI has a clear, accurate view of the entire operational environment.
What is your forecast for the future of IT automation?
I believe we are heading toward a “proactive” era where the concept of “self-service” is replaced by “predictive service” across the entire enterprise. Instead of a user having to navigate a portal to reset a password or request a new cloud environment, the system will recognize the need before the user even voices it, alerting them and guiding them through the process in real-time. We will see the rise of the executive agent, acting as an orchestrator that ensures every automated action aligns perfectly with high-level business goals and strict governance guardrails. By the end of this decade, the most successful organizations will be those that have moved past simple scripts and into a state of “connected workflows,” where a single customer order triggers a governed, autonomous execution through every back-office system, from the database to the cloud infrastructure. It will be a world where IT doesn’t just support the business, but breathes alongside it as an intelligent, invisible partner.
