The Economics of Modern Infrastructure

ABOUT THE EVENT

AI has turned building software into something almost anyone can do. Someone in marketing, finance, or operations can now have an app built, connect it to company data, and deploy it to the cloud - no engineering involved, no security review. Security has lost the single point where it could see what was being created, and now often learns about a new cloud asset only when something goes wrong.

That widening visibility gap is colliding with sharper questions from boards and auditors about how data is governed. When an app built outside any process exposes customer records, it stops being a quiet technical fix and becomes a question the security leader has to answer in the boardroom - no longer about controlling what engineering ships, but about accounting for everything the whole organisation is now putting into the cloud.

DISCUSSION TOPICS

• What are the biggest contributors to Kubernetes spend today?
• Where are organizations seeing over-provisioning and underutilization?
• How are platform teams balancing performance, reliability and cost?
• What approaches are actually driving measurable savings?
• Are FinOps and Platform teams working together effectively?
• How do organizations identify optimization opportunities without impacting production?
• How can engineering teams make infrastructure-cost decisions part of their day-to-day workflow?
• As AI adoption increases, how are organizations thinking about GPU and inference costs?
• What makes inference economics different from traditional cloud workloads?
• How can teams optimize GPU utilization, model serving and inference efficiency?
• Where does the balance sit between latency, throughput, reliability and cost?
• Can Kubernetes optimization and AI inference optimization be managed under one infrastructure-efficiency strategy?
• What metrics should engineering leadership track beyond total cloud spend?
• How do you make optimization continuous rather than a one-time cost-cutting exercise?