Ricardo Castro

I work better if:
- I know where I'm headed
- I have a path to get there
- I work in iterations
- In each iteration I assess progress, evaluate the path, adjust accordingly
- Periodically, I assess if where I'm headed is still the where I should be going

Organizing productive platform teams - Stack Overflow So many platforms feel heavy because they mirror the organization, not the architecture the organization claims to want.

"Organizing productive platform teams" by Adora Nwodo and Peter O'Connor

stackoverflow.blog/2026/03/09/o...

Mastering the OpenTelemetry Transform Processor · Dash0 Learn how to use the OpenTelemetry Transform Processor with practical OTTL examples to safely reshaping logs metrics and traces in the Collector

"Mastering the OpenTelemetry Transform Processor" by Ayooluwa Isaiah

www.dash0.com/guides/opent...

Using Agents to Setup Experiments - ISE Developer Blog When setting up an experiment is complex, agents can automate the process to make it faster and more reliable.

"Using Agents to Setup Experiments" by Peter Lasne

devblogs.microsoft.com/ise/using-ag...

How Data 360 Optimized Kubernetes Scheduling Architecture Explore how Salesforce optimized infrastructure cost at scale by evolving Kubernetes scheduler behavior to eliminate node fragmentation, and much more.

"How Data 360 Optimized Kubernetes Scheduling Architecture, Delivering 13% Cost Savings" by Padma Aradhyula

engineering.salesforce.com/how-data-360...

Revision 165 Articles and updates:

Revision 165 is out!

@koslib.com

#devops #sre #platformengineering

open.substack.com/pub/embracer...

My face after just watching another video about how DevOps is completely different from Platform Engineering and how Platform Engineering is much more than DevOps.

Clearly the author has no idea what they're talking about.

Parse, Don't Guess - Event-Driven.io Event-Driven by Oskar Dudycz

"Parse, Don't Guess" by Oskar Dudycz

event-driven.io/en/parse_don...

Zalando Engineering Blog - Why We Ditched Flink Table API Joins: Cutting State by 75% with DataStream Unions The beauty of a high-level abstraction is that it lets you focus on the

"Why We Ditched Flink Table API Joins: Cutting State by 75% with DataStream Unions" by Maryna Kryvko

engineering.zalando.com/posts/2026/0...

From the RedditEng community on Reddit Explore this post and more from the RedditEng community

"How Reddit Does Threat Detection" by sassyshalimar

www.reddit.com/r/RedditEng/...

Observability Maturity Model - Where Does Your Team Stand? Learn the 5 levels of observability maturity and how to progress through each stage using OpenTelemetry and observability backend. Includes self-assessment table and code examples.

"Observability Maturity Model - Where Does Your Team Stand?" by Saurabh Patil

signoz.io/guides/obser...

If you want Agents to work more autonomously, you need to provide them with the tools to understand if they're doing the right thing.

If not, you're just opening the door for a lot of problems to emerge.

Output != Outcome.

Many OSS projects I follow/contribute to are suffering from very low quality PRs since AI made it easier to produce more output.

This isn't simply an annoyance: it's actually harming those projects due to the sheer volume of stuff being thrown at them.

Using Claude Code for Database Work | Airbyte Learn how to use Claude Code for database work, including querying, schema design, migrations, optimization, and debugging workflows for faster and more reliable development.

"Using Claude Code for Database Work" by Francis Genet

airbyte.com/blog/claude-...

Video Conferencing with Postgres — PlanetScale Stream real time video and audio through PostgreSQL on PlanetScale.

"Video Conferencing with Postgres" by Nick Van Wiggeren

planetscale.com/blog/video-c...

Mount Mayhem at Netflix: Scaling Containers on Modern CPUs Authors: Harshad Sane, Andrew Halaney

"Mount Mayhem at Netflix: Scaling Containers on Modern CPUs" by Harshad Sane and Andrew Halaney

netflixtechblog.com/mount-mayhem...

Beyond the Fix: Building an In-House Terraform Review Agent with Gemini CLI How we turned a CLI tool into a dynamic infrastructure-as-code reviewer using GitLab CI and GitLab MCP.

"Beyond the Fix: Building an In-House Terraform Review Agent with Gemini CLI" by Maksym Kursin

medium.com/@m0ps/beyond...

5 Incident Response Principles for CTOs - Uptime Labs Most organisations don’t truly feel the importance of reliability until they’ve lived through a period of instability - a run of incidents that comes in threes, as the joke goes. That’s usually when…

"5 Incident Response Principles for CTOs" by Joe Mckevitt

uptimelabs.io/articles/inc...

Crossplane & AI: The Case for API-First Infrastructure AI has accelerated coding, but the real bottleneck starts after git push. Infrastructure and operations still rely on fragmented, human workflows. To enable AI agents, platforms must be API-first and…

"Crossplane & AI: The Case for API-First Infrastructure" by Ana Margarita Medina

blog.crossplane.io/crossplane-a...

Revision 164 Articles and updates:

Revision 164 is out!

@koslib.com

#devops #sre #platformengineering

embracerisk.substack.com/p/revision-164

Behind the Fizzy Infrastructure Lead Programmer Kevin McConnell shares the ambitious infrastructure experiment behind Fizzy and the choice to pivot before launch.

"Behind the Fizzy Infrastructure" by Kimberly Rhodes, Fernando Olivares and Kevin McConnell

dev.37signals.com/fizzy-infras...

Inside Platform Engineering with Matt Gowie Infrastructure as code sounds simple until it isn’t. Matt Gowie, founder of IaC consulting firm MasterPoint, joined me on Inside Platform Engineering to share what he’s learned helping organizations…

"Inside Platform Engineering with Matt Gowie

octopus.com/blog/inside-...

You may fool yourself into thinking real state lives in Git.

Git stores your desired state.

The actual state lives in the providers you use (e.g., Cloud Provider).

Continue local sessions from any device with Remote Control - Claude Code Docs Continue a local Claude Code session from your phone, tablet, or any browser using Remote Control. Works with claude.ai/code and the Claude mobile app.

"Continue local sessions from any device with Remote Control"

code.claude.com/docs/en/remo...

Once you peak beneath the surface, things are invariably more nuanced than you thought they were.

How hosted control planes are getting smarter about resource management | Red Hat Developer Learn how hosted control planes reduce costs without compromising performance by using intelligent scaling

"How hosted control planes are getting smarter about resource management" by Ramon Acedo Rodriguez

developers.redhat.com/articles/202...

The fact that, using AI, you can produce more output means that the following are even more important now:

- You need to know where you're going, in order for that output to take you there

- You need easy, cheap and fast ways to validate AI output.

LinkedIn Pulse

"From Observability to Control Loops Part 2: The Risk of Blind Autonomy" by Andrew Mallaband

www.linkedin.com/pulse/from-o...

Troubleshooting Microservices with OpenTelemetry Distributed Tracing Distributed tracing doesn’t just show you what happened. It shows you why things broke. While logs tell you a service returned a 500 error and metrics show latency spiked, only traces reveal the full…

"Troubleshooting Microservices with OpenTelemetry Distributed Tracing"

sematext.com/blog/trouble...

Is your platform team building in the dark? YouTube video by Operationally Intelligent

I had a really interesting conversation with Adam Kinniburgh, on the Operationally Intelligent podcast, about Platform Engineering.

Check it out!

youtu.be/W805caNDJyQ?...