Field notes
Engineering, written down.
A working index of the decisions behind cloud infrastructure, delivery, MLOps, and the software that makes those systems legible.
Cloud infrastructure
Cloud systems are easier to operate when the important decisions are visible before the first resource exists.
DevOps and delivery
A delivery pipeline is a control system: it should make the safe path easy and leave evidence when it says no.
Machine learning operations
A model endpoint is a production system with a second correctness problem: it can be healthy, fast, and wrong.
Security as a delivery property
Security is strongest when it is attached to the normal path of building and shipping, not bolted on as a separate ceremony.
Backend systems
A backend is a set of contracts under load: data shape, failure semantics, latency, and the boundary where trust changes.
Frontend systems
The frontend is a reading instrument: it should make hierarchy, state, and the next useful action obvious.
Architecture notes
Good architecture is a record of trade-offs that another person can inspect, change, and operate.
Tools and working agreements
Tools are useful when they make a decision faster, safer, or easier to revisit.
Applied, not hypothetical
These notes are deliberately opinionated and incomplete. The project records show where the ideas meet real constraints, and the lab records show where they fail.