By Peeyoosh, CEO, Hoonartek
As enterprises embed autonomous decision-making into their core operations, IT services work is undergoing a structural shift — from task execution to decision stewardship. The question is no longer whether this change is coming. It is already here.
For decades, the rhythm of enterprise IT services has followed a familiar pattern. Banks defined credit processes, insurers built claims workflows, telecom operators designed network operations playbooks — and IT services partners executed them with precision and scale. Stability came from predictability. Success meant delivering what was specified.
Agentic AI is beginning to reshape that long-standing contract.
This is not about new tools entering existing workflows. It is about the nature of enterprise work itself changing — from process adherence to autonomous decision-making. As organisations in banking, insurance, and telecom embed AI systems that can observe context, evaluate trade-offs, and act without human prompting, the IT services projects that support them must evolve just as fundamentally.
From Process Automation to Decision Ownership
Most large-scale enterprise IT programmes today are still built around processes — credit underwriting workflows, claims pipelines, customer service escalation paths. The underlying assumption is that if the process is well-defined, the technology will deliver the right outcome. Agentic AI shifts that assumption entirely.
Instead of executing predefined rules, enterprises are now deploying systems that continuously evaluate context, make dynamic trade-offs, and learn from real-world outcomes. A fraud detection model no longer simply flags a transaction — it decides whether to pause it. A collections platform no longer surfaces a list of accounts — it determines which customer engagement strategy to trigger, in real time.
Technology leaders across BFSI and telecom are no longer asking whether systems can be built. Increasingly, they are asking whether those systems can operate within clearly defined business intent and acceptable risk — and whether that clarity can be sustained over time. That represents a structural redefinition of what IT services means.
This transition directly reshapes the type of work IT services teams are asked to deliver — and the capabilities they need to bring to the table.
How Enterprise IT Projects Are Changing
The earliest effects of agentic AI adoption are visible across three familiar areas of enterprise IT services work.
Requirements work is becoming continuous, not ceremonial. In traditional banking and insurance programmes, requirements are gathered, documented, approved, and frozen. Agentic systems operate differently. A credit decision model that continuously adapts to portfolio performance and regulatory shifts cannot be built on a fixed specification.
Business analysts and domain consultants are increasingly shifting toward defining decision boundaries, translating policy intent into machine-interpretable constraints, and refining system behaviour iteratively alongside business stakeholders. The phase-gate delivery model is giving way to something more like ongoing co-governance.
Operations work is evolving from reaction to exception handling. In telecom and enterprise environments, a large portion of IT services effort today goes into monitoring dashboards, investigating alerts, and executing remediation steps. Agentic operations systems are rapidly absorbing this workload — performing autonomous root-cause analysis, executing remediation actions, and preventing recurrence through learning. What remains for IT services teams is more demanding: exception handling, decision audits, trust validation, and designing the escalation and override mechanisms that ensure autonomous systems stay within acceptable boundaries. Operational roles are evolving from reactive support to what can best be described as system stewardship.
Analytics is moving from insight delivery to outcome accountability. Enterprises no longer want dashboards that explain what happened. They want systems that act. Collections platforms dynamically select customer engagement strategies. Churn-prevention systems trigger interventions before a customer even signals intent to leave. Fraud systems pause high-risk transactions before loss occurs. In this environment, data and AI teams are no longer evaluated on model accuracy alone — they are increasingly accountable for the decisions influenced, the actions triggered, and the business outcomes achieved, responsibly and at scale.
New Categories of IT Services Work Are Emerging
As enterprises embed agentic systems into core operations, new forms of IT services work are becoming visible — work that traditional delivery models were never designed for.
Agent behaviour design — shaping how systems reason, prioritise trade-offs, and adapt under uncertainty — is becoming a distinct discipline. Human-in-the-loop governance — defining when and how humans must intervene in autonomous decisions, and building the override mechanisms to support this — is moving from an afterthought to a core engineering requirement. Continuous learning engineering — ensuring systems improve from real-world outcomes without violating compliance or ethical boundaries — is something enterprises are beginning to commission explicitly, even if formal job titles are still evolving.
These are not fringe capabilities. They are rapidly becoming central to what enterprise AI programmes demand — and what IT services partners will be expected to provide.
The Real Talent Challenge Is Job Mismatch, Not Job Loss
Much of the public discourse around AI focuses on displacement. In practice, the more immediate challenge for IT services organisations is job mismatch. Skills built for static workflows, rule-based automation, and one-time system implementations do not translate cleanly into environments where systems adapt continuously, execute autonomously, and require ongoing supervision and governance.
The most valuable IT services professionals in the agentic era will be those who can think in terms of systems and decisions — not tasks. Those who understand that autonomy and accountability are two sides of the same coin.
Those who can translate business intent into governed machine behaviour and revisit that translation as the business evolves. Building this capability at scale is one of the defining workforce challenges the IT services industry faces today.
What Enterprises Will Expect from IT Services Partners
As banks, insurers, and telecom operators mature in their agentic AI journeys, expectations from IT services partners are beginning to shift in a consistent direction. The ask is moving from ‘Can you automate this process?’ to ‘Can you ensure the right decisions happen consistently?’ From ‘Can you deliver the project?’ to ‘Can you help us evolve the system over time?’ These are not incremental changes in scope. They represent a structural shift in the nature of IT services engagement — from project delivery to ongoing decision stewardship.
This evolution has implications for delivery models, commercial structures, talent strategies, and how IT services organisations position their value to clients. The partners who recognise this early — and invest in building the capabilities to support it — will help define the next decade of enterprise technology services.
What Lies Ahead
Agentic AI does not eliminate IT services work. It moves where human effort is applied — and raises the bar for what that effort must deliver. As enterprises adopt systems capable of autonomous decision-making, the work of IT services shifts from execution to judgment, from process adherence to decision oversight, from building systems to shaping and governing their behaviour over time.
The future of IT services will not be defined by how efficiently tasks are completed. It will be defined by how effectively organisations translate business intent into governed autonomy — and how clearly accountability is maintained when systems act at scale. That is the structural shift now unfolding across enterprise technology programmes. And it is already reshaping every engagement, every talent conversation, and every strategic discussion in the IT services industry.
— Peeyoosh is the CEO of Hoonartek. He writes on enterprise technology strategy, AI adoption, and the future of IT services.