the native papersan independent publication by the native web
Want a heads-up when a new paper is out?
Latest Issue · No. 6 · November 2026
One Idea. One Paper. Every Month.
A growing series of concept-focused whitepapers on software architecture, event sourcing, APIs, and AI. One email is all it takes – every new issue, and the full archive from day one.
Every issue is gated. Your email unlocks all of them – past and future.
No. 6 · November 2026
Your Data Team Is Not Too Small
The report is late, and the reflex is to grow the data team. That misses the cause: a central team is preparing data whose meaning belongs to someone else. Data Mesh answers this correctly, by handing data back to the domains that own it. What it assumes, and what most companies never built, is domains. This paper is about that gap, and about what closes it.
Monolith or services gets argued as a matter of taste, and it is a matter of cost. The bill arrives years later, in knowledge no market can supply. Operational complexity can be hired, trained, or bought as a service. The complexity that grows inside your domain cannot. This paper is about that asymmetry, and about the decision it puts on your desk.
An agent writes code faster than your team can review it, and hiring more reviewers does not fix it. That is the real bottleneck of AI-driven development. The way out is to stop inspecting the code and start specifying the behavior it has to produce. This paper is about how that works, and what it costs.
Sovereign cloud offerings promise independence, but the label sits on the wrong level: the law follows the provider's corporate parent, not your data's location. Real sovereignty is the ability to leave, and that is decided in your architecture. This paper is about that decision, and why it is yours to make.
You Don't Even Know Your Most Important Questions Yet
One day someone asks what actually happened, and most systems cannot say: they store only the current state and overwrite the rest. Event Sourcing keeps the events as they happen, so the past stays available for audits, analytics, and the questions you cannot foresee. That makes it a decision for technical leadership.
Artificial intelligence now writes code faster than any human, and often better. That is not a threat to your engineers. It is a shift in what makes them valuable, and it lands squarely on your desk as a leadership decision. This paper is about that shift, and about the talent choices it forces.