Real software
Systems, boundaries, and decisions that must hold up outside the demo.
ExploreNotes from real software
Real-world software development: decisions, systems, and AI without the hype.
I analyse architectures, workflows, and tools through daily work: what works, what breaks, and what still needs to be verified.
decision-log.ts● LIVE
type Evidence = 'observed' | 'inferred' | 'verify';
const decision = {
context: 'production constraints',
signal: measure(realUsage),
tradeoff: document(alternatives),
status: 'observed'
};
// Evidence first. Opinion second.
Focus
Production
Approach
Evidence-led
Scope
Software + AI
Audience
Developers
01 / Introduction
A reference point for software builders.
Here you will find content, tools, and deep dives into the technologies shaping web development.
Whether you are getting started or staying current, the goal is to help you grow with practical technical material.
presentation.mp4Play / 00:33
02 / Topics
Three areas. One craft.
Content for people who have to make technical decisions, not simply learn new tools.
03 / From the notebook
Featured notes
Software systems
The right boundary matters more than the right framework.
Small modules, readable contracts, and declared responsibilities: architecture becomes useful when it makes decisions reversible.
Continue in the NotebookAI workflow
AI accelerates the work you already know how to evaluate.
Technical judgment
A plausible answer is not proof yet.
04 / Latest release
Latest video
🔥2-Next Js Panoramica_
Watch the video05 / Method
Separating facts from the story we tell about facts.
The notebook uses three evidence states to make the degree of certainty behind every technical decision readable.
01
Observed
I start with what I can point to: system behaviour, code, logs, tests, measurements, and stated constraints. Observation must remain distinguishable from interpretation.
02
Inferred
I connect the signals and propose an explanation. An inference can be useful and still provisional: making it explicit lets others challenge, improve, or replace it.
03
To verify
I mark what is missing: an edge case, a production measurement, a primary source, or an experiment. An open question is not a weakness; it is the next verifiable step.
06 / From the field
Editorial principle
The most useful analyses do not give a ready-made answer. They help you ask a better technical question.
A working principle
08 / FAQ
Frequently asked questions
Who are these materials for?
Mid-senior developers, tech leads, and people working on real systems who want to sharpen their technical judgment.
Is this a course platform?
No. It is a personal editorial project made of videos, technical notes, experiments, and analysis.
How do you approach AI content?
With practical tests, explicit limits, and attention to impact on the process, not the novelty of the tool.
