Software Delivery Beyond Code
AI writes code faster than ever — but can you see what it’s doing in production? See how observability, platform engineering and AI are reshaping developer experience across Australia’s largest organisations.
Based on a survey of 154 technology leaders and engineers.
The code got faster. The visibility didn’t keep up.
AI has changed how software gets built. Code is generated in seconds, delivery is faster and more distributed, and expectations have never been higher. But speed has outpaced visibility. Organisations are now shipping more software than they can fully see — and the developers building it are hitting friction their leaders often don’t know exists.
That gap — between what leaders believe is happening and what developers experience every day — is where delivery quietly slows, quality slips, and good engineers start to look elsewhere. This study set out to measure it.
Leaders see the output. Developers experience the work behind it.
run AI-powered applications — but only 9% can see how they behave end to end.
say AI outputs are often inaccurate or incomplete — for lack of production context.
Leaders think developers spend 43% of the week building features. Developers say 24%.
What the study reveals
Where developer time really goes — and why the hidden work of finding context and switching between tools costs more than most leaders realise.
Why 77% of organisations are investing in platform engineering — and the single factor that decides whether it removes friction or adds to it.
How observability is shifting from a post-incident diagnostic into a live input developers rely on while they build.
The AI-specific capabilities most teams are missing — and why organisations are far better at detecting AI issues than resolving them.
Five practical moves for building a software delivery environment that works for developers and leaders alike.
Written for the people who own delivery
For the leaders responsible for how software gets built and delivered — CTOs, VPs and Heads of Engineering, platform and developer-experience leads, and engineering managers at large organisations navigating AI-assisted development.
Independent research, Australian data
The findings draw on the 2026 Ecosystm Developer Experience in Australia Study — an independent survey of 154 technology leaders and engineering practitioners at Australian organisations with more than 500 employees.
Produced by Avocado Consulting in partnership with Dynatrace.
What the press is saying about the study
Australian firms race ahead on AI despite oversight gaps
Why so few organisations can trace how their AI behaves in production, and how maturity drops at each step from detecting an issue to diagnosing and fixing it.
Read on IT BriefBoards are approving AI rollouts they can’t actually oversee, new Australian research finds
What the findings mean for boards: fragmented ownership of AI risk, telemetry that isn’t shaping development, and why governance and observability need to work as one.
Read on Tech Business NewsFind out which side of the gap your organisation is on
Download Software Delivery Beyond Code for the full findings and five practical moves you can act on.
Download the report