All posts

Figuring out why I'm so obsessed with them, from a classical and theory lens.

Stories from the last 20 years of AWS on testing, abstractions, and operations.

Stop shipping bugs! Test changes to avoid breaking prod and clogging your CI/CD pipeline.

Keep agents from wandering off by adding deterministic guardrails.

Corey Quinn has incredible wit. We talk AWS history and DevOps at AWS.

How AI lets us combine the best of every operational model.

The part of DevOps, SRE, Platform Engineering, and Frontline Ops that actually matters.

An article in Infoworld about focusing on quality with coding agents.

Property-based testing and spec-driven development with The Engineering Room.

I had Amazon Quick search my calendar so I could retrace my steps. Mug rescued!

AWS DevOps Agent investigates multiple leads in parallel, and now you can see what it's up to!

Bring your own skills, and see the ones that the agent writes for you as it learns.

Custom skills, learned skills, code indexing, and on-demand SRE tasks.

The AI Economy interviews me about the GA of AWS DevOps and Security Agents.

New York Times Magazine interviewed me about how agents are changing software engineering.

Software Engineering Daily invited me on to talk about Kiro and spec-driven development.

The Stack Overflow Podcast hosted a conversation about AWS history.

A walkthrough of AWS DevOps Agent from its initial launch.

Techstrong TV has me on to talk spec-driven development with Kiro.

A well-produced marketing video where I talk about side quests.

David and Darko Mesaroš talk Kiro and show it in action.

The Grand Unified Theory of Incident Management and how to never reach a dead end.

Demos of new connected telemetry and transaction search features in CloudWatch.

Observability is a means to an end, and this talk is about how to achieve that.

Had some fun with Adobe AfterEffects. Spent hours, and got an 8 second animation.

Demoing auto-clustering of log patterns and generating queries in CloudWatch Logs.

Stories from running incident calls, to deployment and change safety.

A bunch of specific practices to use observability to solve real problems.

Sometimes systems overload just from things slowing down.

Five stories about improving operational culture.

How to make monitoring measure the real customer experience.

How to shed load discriminately.

Lightning talks on distributed systems and ops by AWS' best engineers.

How to do load shedding, dependency isolation, queue management, and ops in serverless.

The temptation to queue work and deal with it later can become undrainable.

The most well-intentioned health checks can lead to even worse outages.

You can't operate a service effectively if you don't instrument it.

Strategies for maintaining predictable, consistent performance in the face of overload.

An overview of DynamoDB and how to think about data modeling.