# Articles

Long-form analyses on systems engineering, product development, decision work, and evidence in complex technical organizations.

The recurring thread: how requirements, architecture, evidence, and collaboration can be connected so teams make better decisions in real projects while staying deliverable.

## Articles
- [Using AI When the Best Models Aren’t Allowed to See the Most Important Data](https://wido-loeffler.com/en/articles/ki-nutzen-wenn-modelle-daten-nicht-sehen-duerfen.md) — How enterprises can use AI when sensitive data cannot go to the strongest cloud models: problem contracts, controlled contexts, and verified outputs.
- [Open Loop Life: Why AI agents suddenly make our open loops visible](https://wido-loeffler.com/en/articles/open-loop-life.md) — How AI agents turn open tasks into dynamic loops and why Decision Debt, Orchestration Tax, and new attention obligations emerge.
- [SysML v2: More precision instead of simplification — why that’s good (and what it costs)](https://wido-loeffler.com/en/articles/sysml-v2.md) — Why SysML v2 enables more precision and testable claims without making the complexity of technical systems simply disappear.
- [Nexus V: A new, decision-centered view of the V-model in regulated industries](https://wido-loeffler.com/en/articles/nexus-v.md) — A decision-centered interpretation of the V-model: traceability as a knowledge graph, arbitration, anchor points, and continuous compliance.