Rob Kohler, Managing Director, Global Infrastructure Services, DXC APJ‑MEA, discusses the complexities of industries operating across a mix of legacy systems, cloud platforms and outsourced services. He explores the emerging need for orchestration across the entire IT estate and examines why AI adoption in mission-critical environments must remain human-led, with accountability for decisions resting firmly with experienced professionals.
Infrastructure Services,
DXC APJ‑MEA
Across Australia, organisations are relying on increasingly complex technology environments to deliver essential services. From government agencies to infrastructure providers, most are now operating across a mix of legacy systems, cloud platforms and outsourced services, often stitched together over years rather than designed as a cohesive whole.
For industries that underpin the economy, such as mining, energy, transport and the public sector, this complexity creates a very real challenge. These are environments where downtime is not tolerated, and where even minor disruptions can have significant operational and reputational consequences.
At the same time, organisations are working within tighter constraints. Persistent skills shortages, ongoing cost pressures and rising regulatory expectations are forcing IT leaders to rethink how operations are managed and sustained. The traditional model – reactive, fragmented and heavily dependent on manual coordination –is becoming increasingly difficult to maintain.
In response, many organisations have invested heavily in automation. While this has improved efficiency in specific areas, it has not addressed the underlying issue: systems are still operating in silos. Automation can accelerate activity, but without coordination, it can just as easily amplify complexity.
What’s emerging instead is a need for orchestration. This means connecting systems, data and workflows across the entire IT estate and managing them as a unified whole. Rather than adding more tools, enterprises need a clear, enterprise-wide understanding of what is happening across the environment at any given time. The value of this is the ability to see how everything fits together – how performance, risk and cost interact in real time and how to act on that insight with greater confidence.
I’ve seen this challenge play out consistently in conversations with Australian organisations. Many organisations don’t lack technology; they lack a way to bring together complex, multi-vendor environments into a single, reliable operational view.
This becomes even more relevant as AI moves from pilots to practical deployment. Across Australia, there is a clear shift underway, from broad AI ambition to a much stronger focus on implementation. Organisations are now thinking more seriously about how AI operates within existing environments and how it is governed.
In IT operations, AI is already delivering tangible benefits. It can identify patterns across vast datasets, predict potential issues before they escalate and recommend actions in real time. Used well, this has the potential to significantly improve resilience and reduce operational risk.
However, the introduction of AI also raises important questions, particularly in mission-critical environments: How are decisions made; how are they validated; and who retains accountability when something goes wrong?
This is where a human-led approach becomes essential. Organisations are increasingly looking to combine AI with human expertise, using technology to handle scale and complexity while ensuring experienced professionals remain responsible for judgement and decision-making.
In practice, this means AI operates as an enabler, not a replacement. It surfaces insights, highlights risks and suggests actions, but does so within a framework that is transparent, governed and accountable. This balance is critical in sectors where trust and reliability are non-negotiable.
Models need to reflect this thinking in action. AI agents should operate alongside human experts, continuously interpreting signals and identifying potential issues, while decision-making should remain grounded in context and experience. This more pragmatic approach to AI recognises both its potential and its limits.
For Australian organisations, this shift also supports a broader change in how IT operations are measured. There is growing emphasis on outcomes, connecting technology performance to business value, service delivery and operational resilience, while reducing reliance on input-based measures like headcount or service levels.
Where we have organisations managing the combined impact of skills shortages, cost constraints and increasing regulatory scrutiny across environments that cannot afford disruption, improving operations means strengthening governance, coordination and accountability across the existing estate.
Greater transparency and a more connected view of operations play a key role here, enabling organisations to manage performance, risk and cost more cohesively while maintaining clear lines of accountability.
There is strong digital ambition across both public and private sectors in Australia, reinforced by the recently announced federal budget which places AI adoption, digital capability and cyber resilience at the centre of productivity and economic reform. The organisations that will benefit most are those that get this balance right: equipping experts with AI that works at scale, while ensuring humans remain firmly in control of the decisions that matter most.
