Sydney Engineering and DevOps Summit 2027

Beat the rush and join us early for complimentary barista-made coffee and breakfast.

AI is moving beyond simply assisting developers. AI agents are beginning to take on parts of the software delivery lifecycle themselves, from writing and testing code to managing deployments and infrastructure.

This keynote explores what happens when AI becomes an active participant in software delivery, how cloud-native environments are enabling this shift, and what engineering leaders need to do to prepare their teams, processes, and platforms for a more autonomous future.

• How software delivery changes when AI agents actively participate in development, testing, and deployment.
• Where cloud-native technologies are helping teams adopt agentic AI — and where they may be adding complexity.
• The skills, processes, and leadership capabilities engineering teams need to build for AI-driven software delivery.

Agents are provisioning infrastructure, pushing deployments, and triaging incidents with less human oversight than most leadership teams realise.

This keynote asks the question nobody wants to answer in the boardroom: who's actually accountable when an agent breaks something at 2am?

• Where agentic DevOps is delivering real gains, and where it's smoke and mirrors.
• What changes when agents can act on infrastructure, not just suggest code.
• The rollback and oversight mechanisms most teams don't have yet, and badly need.

A meaningful chunk of the industry chased microservices because everyone else was doing it, then quietly paid for it in operational overhead.

This panel puts the sacred cow on the table and asks who actually needed to split their monolith in the first place.

• Why teams are walking microservices back into modular monoliths, and not quietly.
• A blunt framework for architecture choice based on your team, not your ego.
• What's brutally hard to undo once you've committed to the wrong one.

A team ripped out the CI/CD steps that were slowing them down more than protecting them, and release velocity jumped.

This case study covers what they cut, what they were terrified to cut, and what actually happened when they did.

• Which of their "essential" pipeline steps turned out to be dead weight.
• What they automated first for the fastest visible win.
• The real change in deployment frequency and failure rate afterward.

A hands-on, interactive session working through a real engineering and DevOps scenario as a room. Details announced soon.

Platform engineering has an adoption problem nobody likes to admit: teams build an internal developer platform, and developers quietly route around it.

This case study covers the failed first attempt, what was actually wrong with it, and what turned adoption around.

• Why the first version of their platform got ignored.
• What changed to get developers using it instead of avoiding it.
• The velocity and onboarding gains once adoption actually stuck.

Kubernetes at enterprise scale stopped being a technology decision and started being a cost and reliability crisis.

This case study covers where it broke down, what autoscaling actually fixed versus what it made worse, and what they'd never do again.

• Where their Kubernetes environment quietly became unmanageable.
• Their real approach to autoscaling and multi-cluster sprawl.
• What they'd refuse to repeat if starting over.

Perimeter security was never built for the volume of service-to-service traffic modern architectures generate, and most organisations know it.

This panel calls out where service mesh security is genuinely solving that, and where it's a checkbox nobody's tested.

• Why perimeter thinking fails completely for service-to-service traffic.
• Where service mesh is actually enforcing zero-trust, and where it's theatre.
• The gaps in service mesh security implementations nobody wants to admit to.

Small-group, discussion-based sessions where you'll work through real engineering and DevOps challenges with peers in similar roles. Roundtable topics will be announced soon.

Put your knowledge to the test in this fast-paced quiz covering real-world trivia, key concepts, and emerging trends. Compete for bragging rights — and a voucher — as the top scorer takes the crown.

AI workloads are now blowing cloud budgets industry-wide, and most FinOps functions found out the hard way.

This case study is an honest account of discovering runaway AI spend, and what actually stopped the bleeding, not just a dashboard nobody checks.

• How AI and cloud costs grew invisibly before anyone raised the alarm.
• The FinOps changes that actually moved the number, not just visibility theatre.
• How they kept cost control from becoming an engineering blocker.

Serverless was pitched as the end of infrastructure headaches, and for some workloads it delivers exactly that. For everyone else, it quietly created new cost and latency problems they're still cleaning up.

This keynote gives the unsponsored, warts-and-all verdict.

• The workloads where serverless is a genuine, defensible win.
• Where default serverless adoption has quietly cost teams money and performance.
• A blunt framework for choosing serverless, containers, or neither.

Most teams have monitoring dashboards. Very few have real observability, the ability to actually explain a failure within the hour.

This interactive session puts the room's honesty to the test with live voting.

1. Can your team explain why a production incident happened within the first hour, most of the time, or rarely?
2. What's your biggest observability blind spot: logs, metrics, traces, or AI/ML workloads?
3. Has an incident reached customers before your monitoring caught it in the last six months?
4. Is your observability tooling spend outpacing your infrastructure spend?
5. Where's your next investment going: better tooling, better instrumentation, or better on-call process?

Unwind with your peers for a couple of drinks on us!

Beat the rush and join us early for complimentary barista-made coffee and breakfast.

AI is moving beyond simply assisting developers. AI agents are beginning to take on parts of the software delivery lifecycle themselves, from writing and testing code to managing deployments and infrastructure.

This keynote explores what happens when AI becomes an active participant in software delivery, how cloud-native environments are enabling this shift, and what engineering leaders need to do to prepare their teams, processes, and platforms for a more autonomous future.

• How software delivery changes when AI agents actively participate in development, testing, and deployment.
• Where cloud-native technologies are helping teams adopt agentic AI — and where they may be adding complexity.
• The skills, processes, and leadership capabilities engineering teams need to build for AI-driven software delivery.

Agents are provisioning infrastructure, pushing deployments, and triaging incidents with less human oversight than most leadership teams realise.

This keynote asks the question nobody wants to answer in the boardroom: who's actually accountable when an agent breaks something at 2am?

• Where agentic DevOps is delivering real gains, and where it's smoke and mirrors.
• What changes when agents can act on infrastructure, not just suggest code.
• The rollback and oversight mechanisms most teams don't have yet, and badly need.

A meaningful chunk of the industry chased microservices because everyone else was doing it, then quietly paid for it in operational overhead.

This panel puts the sacred cow on the table and asks who actually needed to split their monolith in the first place.

• Why teams are walking microservices back into modular monoliths, and not quietly.
• A blunt framework for architecture choice based on your team, not your ego.
• What's brutally hard to undo once you've committed to the wrong one.

A team ripped out the CI/CD steps that were slowing them down more than protecting them, and release velocity jumped.

This case study covers what they cut, what they were terrified to cut, and what actually happened when they did.

• Which of their "essential" pipeline steps turned out to be dead weight.
• What they automated first for the fastest visible win.
• The real change in deployment frequency and failure rate afterward.

A hands-on, interactive session working through a real engineering and DevOps scenario as a room. Details announced soon.

Platform engineering has an adoption problem nobody likes to admit: teams build an internal developer platform, and developers quietly route around it.

This case study covers the failed first attempt, what was actually wrong with it, and what turned adoption around.

• Why the first version of their platform got ignored.
• What changed to get developers using it instead of avoiding it.
• The velocity and onboarding gains once adoption actually stuck.

Kubernetes at enterprise scale stopped being a technology decision and started being a cost and reliability crisis.

This case study covers where it broke down, what autoscaling actually fixed versus what it made worse, and what they'd never do again.

• Where their Kubernetes environment quietly became unmanageable.
• Their real approach to autoscaling and multi-cluster sprawl.
• What they'd refuse to repeat if starting over.

Perimeter security was never built for the volume of service-to-service traffic modern architectures generate, and most organisations know it.

This panel calls out where service mesh security is genuinely solving that, and where it's a checkbox nobody's tested.

• Why perimeter thinking fails completely for service-to-service traffic.
• Where service mesh is actually enforcing zero-trust, and where it's theatre.
• The gaps in service mesh security implementations nobody wants to admit to.

Small-group, discussion-based sessions where you'll work through real engineering and DevOps challenges with peers in similar roles. Roundtable topics will be announced soon.

Put your knowledge to the test in this fast-paced quiz covering real-world trivia, key concepts, and emerging trends. Compete for bragging rights — and a voucher — as the top scorer takes the crown.

AI workloads are now blowing cloud budgets industry-wide, and most FinOps functions found out the hard way.

This case study is an honest account of discovering runaway AI spend, and what actually stopped the bleeding, not just a dashboard nobody checks.

• How AI and cloud costs grew invisibly before anyone raised the alarm.
• The FinOps changes that actually moved the number, not just visibility theatre.
• How they kept cost control from becoming an engineering blocker.

Serverless was pitched as the end of infrastructure headaches, and for some workloads it delivers exactly that. For everyone else, it quietly created new cost and latency problems they're still cleaning up.

This keynote gives the unsponsored, warts-and-all verdict.

• The workloads where serverless is a genuine, defensible win.
• Where default serverless adoption has quietly cost teams money and performance.
• A blunt framework for choosing serverless, containers, or neither.

Most teams have monitoring dashboards. Very few have real observability, the ability to actually explain a failure within the hour.

This interactive session puts the room's honesty to the test with live voting.

1. Can your team explain why a production incident happened within the first hour, most of the time, or rarely?
2. What's your biggest observability blind spot: logs, metrics, traces, or AI/ML workloads?
3. Has an incident reached customers before your monitoring caught it in the last six months?
4. Is your observability tooling spend outpacing your infrastructure spend?
5. Where's your next investment going: better tooling, better instrumentation, or better on-call process?

Unwind with your peers for a couple of drinks on us!