THE QUESTIONS CHANGED.THE ENGINEERING GOT DEEPER.
Every stage gave me a harder question to solve.
Curiosity
What ChangedI stopped only using technology and became curious about how it worked underneath the interface.
Question Carried ForwardHow do things actually work underneath the interface?
Building
What ChangedSoftware stopped being something I used and became something I could build.
Question Carried ForwardHow do I build reliable software?
Systems
What ChangedI started thinking beyond individual applications to the systems that host, route, and store them.
Question Carried ForwardHow do systems behave underneath the software?
Operations
What ChangedI learned that writing software is only half the job. Someone has to make it deployable, observable, and recoverable.
Question Carried ForwardHow do I make software deployable and repeatable?
Reliability
What ChangedReliability became an engineering problem rather than an availability percentage.
Question Carried ForwardWho owns reliability once the application is deployed?
Abstraction
What ChangedI realized the bottleneck wasn't just infrastructure—it was the cognitive load on developers. I needed to build abstractions.
Question Carried ForwardHow do I make good infrastructure reusable for other engineers?
Controlled Autonomy
The bottleneck shifted from human execution speed to human reasoning speed during incidents. AI became an operational primitive.
What operational decisions can machines reason about safely?
WHAT I BELIEVE NOW
Build before you abstract.
Measure before you optimize.
Design for failure.
Automate the toil, not the responsibility.
Give systems power only where you can control their failure modes.