How an AI works when you talk to it

Wed, 24 Dec 2025 00:00:00 -0300

Unpacking large language models

When you type a message to ChatGPT or Claude, what happens on the other side is simpler than it looks—but it runs at a scale that is hard to picture.

The essence: a digital football commentator

Picture this: World Cup final, Brazil vs Argentina, last minutes of the match. Galvão Bueno is calling the game, the ball reaches the Brazilian striker, he beats one defender, then another… and suddenly the audio cuts out. The broadcast keeps rolling, but the commentator’s voice is gone.

Jenkins CasC: high-performance pipelines with GitOps

Fri, 13 Jun 2025 00:00:00 -0300

The Need to Evolve: Challenges of “Basic” Jenkins

Jenkins is one of the most powerful and popular automation servers in the world. It drives countless CI/CD pipelines that build, test, and deploy applications. Managed only through the web UI — what many call “ClickOps” — it shows its weak spots.

ClickOps creates problems that hurt agility and reliability on any project, especially in complex MLOps environments:

No versioning or audit trail: Who changed a job config? Why? Without version control those questions are almost impossible to answer. Rolling back a bad change becomes a manual treasure hunt.
Inconsistency and hard reproducibility: Cloning a Jenkins environment for development, staging, or disaster recovery is painful and error-prone. Small config drifts between environments cause unexpected failures.
Painful scale and maintenance: Managing dozens or hundreds of jobs by hand does not scale. Maintenance becomes a bottleneck, and consistency across projects stays out of reach.

To get the reliability, speed, scale, and security modern projects need, you have to go beyond the basics.

Mlops on Rafael Zago

How an AI works when you talk to it

Unpacking large language models

The essence: a digital football commentator

Jenkins CasC: high-performance pipelines with GitOps

The Need to Evolve: Challenges of “Basic” Jenkins