Steve McGhee, Google, and Michael Rembetsy, Bloomberg
Architecture is the art of choosing your tradeoffs. Do you accept the "vendor lock-in" to move faster, or do you drown in the complexity of a "cloud-agnostic" abstraction layer? Do you chase five-nines availability across regions, or do you prioritize data consistency? In this session, we’ll create a forum for open discussion on topics such as Multi-Cloud strategies, the reality of Hybrid/On-Prem, managing critical vendor dependencies, and the unique constraints of highly regulated environments. Join us to compare notes on what works, what failed, and the hidden costs of complexity.
Steve loves to help Google Cloud customers build reliable systems with sustainable teams. He was a Google SRE for over a decade, then a customer, then a Reliability Advocate. He recently returned to his roots in SRE as a member of the Google Cloud Incident Response Core Team. He is also a host of the Prodcast, Google’s podcast on SRE and Production Software.
Michael is the Global Head of Network Engineering and Operations at Bloomberg. His teams are responsible for everything from physical hardware to SREs who focus on global connectivity for customers. Before Bloomberg, Michael was the VP of Infrastructure at Etsy. Michael has spoken around the world, participated in various industry program committees, and is a past conference program chair for LISA and SREcon. Additionally, he has served on advisory boards throughout his career.
author = {Steve McGee and Michael Rembetsy},
title = {Architecture Decisions and Tradeoffs},
year = {2026},
address = {Seattle, WA},
publisher = {USENIX Association},
month = mar
}