Your engineering practice as an agentic pipeline

Last week we had a message from a user asking about a Safari bug. The bug has been lingering for a while. I've been experimenting with an agentic pipeline and decided it was time to test the next level. I gave it the Safari bug and it shipped a PR. The same week it fixed two other bugs that the team did not have time for.

The death of software shipping has been declared more than once. It will be a while before it actually happens, but the process around it is quietly being automated. Agentic DevOps, engineering as infrastructure, lets teams put their way of working into a pipeline, and let's that pipeline carry the work.

Delivering complete features are out of reach for AI for now. All those annoying little bugs the team has no time for or doesn't want to deal with? With clear context AI can absolutely play a role there. Not through vibe coding, but through an agentic pipeline that lifts the quality of what AI produces: tests and linters the agent cannot bypass, review agents to catch what those let through.

It allows humans to focus on the reason software is being built in the first place. Larger features, architectural decisions, stakeholder conversations, the trade-offs that need a human in the room. The repetitive parts of shipping are moving into the pipeline.

There is no one true way to do this, everyone is figuring it out as they go. Experimentation is key. Teams already building an agentic DevOps pipeline are learning what works for them. It's about moving your engineering practice out of engineers' heads and coding guidelines documents and into a pipeline that can pick up work. One step at a time.