This article is part of the How Do YOU CM2? blog series in collaboration with the Institute for Process Excellence (IpX). Although I receive compensation for writing this series, I stand behind its content. I will continue to create and publish high-quality articles that I can fully endorse. Enjoy this new series, and please share your thoughts!
Your change review board approved a design change six months ago. Today, a similar problem has surfaced in a different product line. Nobody remembers why the original solution was chosen over two viable alternatives. The decision was documented. The rationale was not.
This is not a documentation problem. It is a knowledge architecture problem. Research on design rationale has shown for decades that capturing the “why” behind decisions is critical for reuse and learning. Yet it rarely happens, because documenting rationale is intrusive, time-consuming, and disconnected from how engineers actually work.
The real cost is compounding. New hires spend roughly 200 hours working inefficiently because contextual knowledge was never captured. Boeing’s experience in the 2010s demonstrated what happens when institutional memory erodes: integration failures on the 787 traced back to process knowledge that existed in people’s heads but never made it into retrievable records.
This is where AI scaffolding changes the equation. Not AI generating rationale after the fact. AI captures it as a byproduct of the work itself.
In an AI-assisted impact analysis, the engineer and the AI walk through a product structure together, level by level. At each step, the AI surfaces a dependency and explains why it matters. The engineer confirms, rejects, or redirects. That interactive exchange is itself the rationale record. The “why” is captured not because someone stopped to write it down, but because the process required articulating it.
CM2’s closed-loop change process provides the structure that makes this work. The change objects, the baselines, the impact matrices, these are not just governance artifacts. They are the scaffolding that gives AI a consistent framework to interact with. Without that structure, you get a conversation. With it, you get a traceable decision record.
The INCOSE Systems Engineering Handbook states that decisions should be documented using digital engineering artifacts, including the analysis, decisions, and rationale for historical traceability and future decisions. CM2 operationalizes this by defining exactly where in the change process those decisions live and who owns them.
The uncomfortable question: if your organization cannot reconstruct why a decision was made six months ago, what makes you confident the next decision will be any better informed?