Following our quick coverage of Day 2, I’m excited to share the key highlights from Day 3 of SRECon 2025!
If you missed our previous updates, all the important moments are still available for your review.


Now, let's jump into today's most significant developments and memorable sessions.
Highlights from Day 3
Here’s a quick look at the Day 3 sessions that sparked some serious conversation:
Tech Debt
Yvonne and Michael Rembetsy tackled the often-overlooked relationship between tech debt and SRE work. They shared how SRE teams get caught managing not just their technical shortcuts, but also deal with the fallout from debt across the services they support.
If you've ever felt the pain of maintaining systems built on hasty decisions, this talk was a must-see for practical insights on surviving—and reducing—technical debt in production environments.
Production Engineering When Trading Billions of Dollars a Day
Pedro shared an insider's look at what it takes to run trading systems handling billions of dollars daily at Jane Street. He revealed how production engineering works when every millisecond and message directly impacts profits and losses.
Drawing from his seven years of experience building and monitoring Jane Street's trading systems, Pedro offered rare insights into both the daily operations and those heart-stopping moments when things go wrong in high-stakes financial technology.
This talk provided a fascinating glimpse into an environment where reliability isn't just important—it's worth billions.
Observability
Daria tackled the complete observability toolkit—logs, metrics, and distributed traces—while challenging how we measure success in monitoring programs.
As Azure's Principal SRE in Observability Engineering, she combined her mathematics and AI background with global experience from Moscow to the Pacific Northwest to deliver practical insights. Outside tech, she brings an unusual passion for opera to her work improving reliability and on-call experiences.
Daria's discussion offered clear frameworks for evaluating observability effectiveness beyond just collecting data.
From HAR to OpenTelemetry Trace: Redefining Browser Observability
Antonio demonstrated how to transform HTTP Archive (HAR) files into OpenTelemetry traces, creating a powerful new approach to browser observability. The Cisco ThousandEyes Tech Lead revealed his method for converting page load requests into OpenTelemetry-compliant spans that can be streamed to tools like Jaeger or any backend via the OpenTelemetry collector.
If you've been struggling to extract meaningful insights from browser performance data, Antonio's practical architecture opened up new possibilities for deeper web application monitoring beyond traditional methods.
One Million Builds per Year, Only One Page - Operating Internal Services Without Heroics
Cail revealed how a small team at Octopus Deploy managed over one million builds annually with just a single after-hours page.
Drawing from his diverse background in performing arts, film, and software ops, he shared both technical and social approaches that eliminated the need for heroics. Surprisingly, he also explored the unexpected downsides when systems become too stable.
If you're tired of middle-of-the-night alerts, Cail's practical experience offers a roadmap to reliability without burnout.
And that's a wrap for SRECon Americas 2025!

It was great to connect with so many of you and learn about your observability tricks and oops moments. Huge thanks to the organizers, speakers, and sponsors for an unforgettable event!