open-ai-transformation-maturity-model

Home · Getting Started · Model · Assessment Hub · Roadmap · References


Pillar 5 — Codebase Readiness & Brownfield Extractability

License: CC BY-SA 4.0
https://creativecommons.org/licenses/by-sa/4.0/

This pillar assesses whether existing systems can be understood, modified, and evolved reliably, including the ability to extract behavioral knowledge from legacy codebases.

👉 Back to Assessment Framework
👉 Back to Unified Scoring System


Why This Pillar Matters

Most organizations operate complex legacy systems whose behavior is only partially documented or understood.

AI-driven transformation requires the ability to interpret existing code, derive specifications from observed behavior, and safely modify systems without introducing regressions.

Without sufficient codebase readiness, transformation efforts stall at surface-level productivity improvements.


What This Pillar Evaluates

This pillar evaluates whether the organization can:


Typical Assessment Questions

Examples of criteria within this pillar include:


Evidence to Look For

Useful evidence may include:


Low-Maturity Pattern

At low maturity, organizations tend to rely on:

Such systems resist both manual and AI-assisted evolution.


High-Maturity Pattern

At high maturity, organizations tend to exhibit:

Legacy systems become assets rather than constraints.


Common Failure Modes

Knowledge Silos

Critical system understanding is limited to a few individuals.

Architectural Erosion

System structure has drifted from original design without updated documentation.

Change Aversion

Teams avoid modifying legacy components due to unpredictable risk.


Relationship to Other Pillars

This pillar strongly influences:


Scoring Interpretation

Score 0

Legacy systems are poorly understood and difficult to modify safely.

Score 1

Partial understanding exists but is inconsistent and fragile.

Score 2

Key systems can be analyzed and modified with moderate confidence.

Score 3

Architecture and behavior are well understood across major systems.

Score 4

Systems are highly transparent, analyzable, and amenable to automated reasoning and evolution.


Relevance by Level


↑ Back to top · Home