Comparison guide
Prototype software learns. Production software carries obligation.
RDD is useful only if the boundary stays visible. Build to Learn can move quickly because it is provisional. Build to Deliver begins when evidence is strong enough to govern production work.
Prototype versus production
Built to expose disagreement, clarify rules, test workflow assumptions, and produce evidence.
Built to be secure, maintainable, observable, accessible, scalable, integrated, tested, and governed.
The boundary RDD protects
The failure mode is prototype theater: a convincing demo gets treated as proof that delivery will be simple. RDD rejects that move.
The Freeze Gate asks what the prototype proved, what it did not prove, what risks remain, and which Reality Checks must happen before production commitment.
When a prototype has done its job
A prototype has done its job when it makes a decision clearer. It may confirm behavior, reveal an exception, remove scope, expose a policy conflict, or show that the team is not ready to commit.
The artifact worth carrying forward is not the disposable build. It is the evidence record produced by contact with reality.
Prototype risk checklist
- Audience riskWere the right users and stakeholders involved, or only internal reviewers?
- Evidence riskDid the prototype produce accepted behavior, rejected alternatives, and open risks?
- Architecture riskHas anyone tested whether the behavior fits system boundaries, data flows, and integration constraints?
- Security riskHave roles, permissions, audit trails, privacy, and failure modes been challenged?
- Operations riskDoes the organization know who owns the process after release and how exceptions are handled?
Prototype to production handoff
The responsible handoff is not "make the prototype real." The responsible handoff is a delivery baseline that states what the prototype proved, what it did not prove, and which Reality Checks must happen before release.
This distinction matters more in the AI era because teams can now produce convincing candidate software quickly. RDD protects the organization from confusing plausibility with readiness.
FAQs
Can prototype code ever become production code?
Sometimes, but only after it meets production standards. RDD does not assume prototype code is disposable or reusable; it asks what evidence should survive.
What makes production software different?
Production software carries obligations around security, reliability, maintainability, accessibility, observability, data integrity, support, and governance.
Why does RDD use candidate reality instead of prototype?
Candidate reality emphasizes the purpose: a provisional working interpretation used to discover reality before engineering commitment.