A perceptible change has become apparent in how software is built. Late last year, generative AI surged in practical capability, prompting many technology leaders to rethink traditional development workflows. Tools that once played a minor supporting role in coding have become more important. For example, Y Combinator’s Winter 2025 batch saw a quarter of startups with 95% of their code generated by AI, while organizations are broadly reporting developer productivity gains of 20-50% when using AI.
While Australian businesses are seeing improved developer productivity and cost savings as AI absorbs more coding work, coding still only accounts for a small part of software delivery. Creating efficiencies in coding often exposes inefficiencies elsewhere in the development cycle, such as in reviews, testing, security, deployment, or operations.
This is known as the AI paradox, where the issue is not a lack of AI tools but a lack of connection and orchestration between tools. Research shows that roughly two-thirds of DevSecOps teams in Australia use more than five AI and security tools. Organisations need to move their workloads from fragmentation to flow, and reconsider how quality and security span the entire software development lifecycle.
Fragmentation Hinders Software Development
Fragmentation, in many forms, is hindering organisations’ ability to effectively realise value from AI. Some ways that we are seeing this in Australia:
AI tooling: Most enterprises built their software delivery capability tool by tool over the past decade. Each tool now comes with its own AI agent. Developers use one AI for coding, another for security analysis, and another for CI/CD troubleshooting. The problem is that they don’t work together.
AI context: Without a unified data model, each agent operates in its own silo, lacking context about the broader project. Requirements, code history, security implications, deployment constraints, and operational feedback remain disconnected across systems, forcing teams to manually bridge these gaps.
Trust in AI: Even with great AI tooling, trust isn’t a switch one flips. Some developers let AI generate entire modules; others won’t accept a single suggestion without rewriting it. Neither extreme is wrong. Without consistent verification and validation processes, it’s not clear which tasks are well-suited to AI, given quality and risk, and what level of human approval is required.
AI governance: There is a growing need for data residency, and no single deployment model will suffice. Additionally, new AI laws are driving urgent governance requirements to identify and record AI use across both approved and shadow tools. Regulators and industry bodies are also demanding more “prove it” controls. All of which requires a fresh look at AI security and governance.
AI budget: Finance teams see the growing AI “line item” across infrastructure investments and different software tools that every team is buying. They are rightfully asking everyone to be pragmatic, asking for clear usage telemetry, cost controls, and return on investment before pressing further.
The need for a unified architecture
Unified architecture is the best antidote to fragmentation in software development. This replaces sequential stages with continuous execution, where AI agents work within the loop while humans orchestrate.
Organisations need platforms spanning the entire lifecycle, from planning through operations. When agents share a common execution environment, the deployment agent instantly accesses code changes, the security agent automatically triggers remediation, and the performance agent directly informs the architecture. Context persists throughout rather than being lost at handoffs.
For example, at Thales, fragmentation meant teams were completely isolated from one another. Moving to a unified platform transformed their environment, enabling better communication and coordination among their diverse teams across multiple locations.
Additionally, intelligent orchestration requires connecting relationships between code, requirements, tests, security findings, deployments, and metrics throughout the entire organisation. This organisational memory lets agents see full context: who requested a feature and why, what constraints apply, what similar implementations exist, and how changes impact downstream systems. Service catalogs with ownership tracking synthesise developer experience and security metrics to detect drift. When merge request cycle times spike or change-failure rates rise, the system automatically triggers responses. The data model evolves continuously, learning patterns that make every agent smarter.
Also, teams need customisable autonomy to define which context agents rely on, which workflows to streamline, and which compliance rules to enforce. Low-risk changes proceed autonomously. Medium-risk changes trigger review workflows. High-risk changes require explicit approval. Agents can integrate across the enterprise toolchain, pulling context from Jira, PagerDuty, Confluence, and Snowflake, while the unified platform provides orchestration.
Compliance must be embedded throughout with AI threat modeling, automated supply chain security, secrets detection, and comprehensive AI governance. Policy gates enforce rules automatically. Audit trails capture every agent decision. Shadow-agent detection identifies unapproved tools. Continuous compliance monitoring with exportable evidence packs demonstrates governance to regulators. Teams define policies once. The platform enforces them consistently. For example, Singapore-headquartered Agoda leveraged a unified platform to bring consistency to governance, compliance, and security audits across its organisation.
Lastly, organisations need deployment options (SaaS, dedicated instances, self-managed) for local and cloud-hosted models. Transparent usage-based pricing should align costs with value, with visibility into token spend and team-level budget controls. A marketplace approach lets teams choose optimal models for each task rather than paying for bundled capabilities they don’t need.
Future-proofing software development
In Australia, organisations are at a critical inflection point. Most have now moved beyond piloting AI and are seeing real results. The AI paradox is not a short-term challenge; it’s a foundational one. If AI is treated only as a coding accelerator, bottlenecks will grow elsewhere in the lifecycle, and in fast-moving digital economies like Australia’s, that gap will widen quickly.
The longer that AI remains fragmented within an organisation, the more it compounds technical debt, integration complexity, and organisational drag. Now is the time to combine platform consolidation with intelligent orchestration to move faster, focus more on innovation, and change how software gets built.
Click Here For The Original Source.
