Limits of Traditional Enterprise Architecture
TOGAF, Zachman, and classic approaches were designed for stable environments
- Predictable technologies
- Long change cycles
- Deterministic systems
Overly rigid and document-driven
- Focus on artifacts and deliverables
- Low impact on actual execution
- Hard to adapt to rapid change
Centered on applications and infrastructure
- Design oriented to technical systems
- Infrastructure as the central axis
- Poorly aligned with cloud and serverless
No native model for AI and agents
- AI treated as an add-on
- No agents or autonomous decisions
- Lack of an architectural vision for AI
Limited for modern decision-making
- Static processes
- No continuous learning
Key message: Traditional Enterprise Architecture was created for a stable world. Modern organizations operate in a dynamic, intelligent, and automated world. Kotusev documents that, for almost half a century, consultancies and gurus made fortunes selling the same enterprise architecture frameworks as “best practices”, frameworks that demand unreasonable effort to produce and maintain documentation that is overly complex, often outdated, and ends up shelved in “ivory towers”. At best, these frameworks are useless; at worst, harmful [1].
Benefits of ILEA™
| Key Area | Without Enterprise Architecture | With Intelligent & Lean Enterprise Architecture (ILEA) |
|---|---|---|
| Solution Design | Monolithic, fragile solutions that are hard to scale and maintain | Modular, composable, scalable, and reusable architectures |
| Business Alignment | Tactical decisions disconnected from strategy | Architecture guides strategic decisions and continuous execution |
| AI and Agent Integration | Ad hoc use of AI with security risks and low impact | AI and agents governed as reusable structural capabilities |
| Cloud Readiness | Inefficient migrations or limited use of the cloud | Efficient, optimized cloud-native architectures based on PaaS |
| Security and Compliance | Reactive, costly security added at the end | Security embedded from the design stage (security by design & lean) |
| Agility and Evolution | High technical debt and slow change | Continuous evolution at low cost and with high adaptability |
Key message: Without architecture, organizations react. With ILEA, they decide and evolve with intent.
Principles
Artificial Intelligence as a cross-cutting meta-domain
Artificial Intelligence is neither an isolated component nor an additional layer, but a meta-domain that cuts across and strengthens all domains, enabling augmented decisions, automation, and continuous learning.
Agents as operational units
Intelligent agents become first-class architectural actors, capable of executing tasks, orchestrating processes, and making decisions within governed boundaries.
Architecture oriented to value, not infrastructure
Architecture is designed to maximize business value, not to manage infrastructure. Technical complexity is abstracted through managed and serverless platforms.
Adaptability by design
The model assumes constant change as the norm. Architectures must learn, adapt, and evolve without frequent structural redesigns.
Simplicity and a lean approach
Priority is given to eliminating redundancies, unnecessary dependencies, and over-design, keeping only what is essential to operate, scale, and evolve.
Integrated and traceable governance
Automation, AI, and agents operate under clear principles of security, control, and traceability, ensuring trust, compliance, and accountability.
Key message: Enterprise Architecture stops being a documentation exercise and becomes a living, intelligent, value-oriented system.
How ILEA Works
The model operates as an active system
- Domains do not work in isolation but as an interconnected system.
- Artificial Intelligence enables informed, repeatable, and scalable decisions, integrating business context, data, and rules to turn the architecture into a decision system.
- Decisions progressively shift from manual to augmented and automated.
- The organization evolves its way of operating more efficiently, without constant redesigns of the architecture.
Lean as a cross-cutting principle
- Fewer unnecessary layers
- Fewer duplicated controls
- Less dependence on infrastructure
- More focus on value, process improvement and automation, speed, efficiency, and learning.
Continuous value cycle
- Business defines objectives, value, and priority processes
- Data provides semantics and reliable context
- Agents execute and orchestrate decisions and actions
- Applications & Integration enable end-to-end capabilities
- Security ensures confidentiality, integrity, availability, control, and compliance
- Platform scales, modernizes, and optimizes operations
- Artificial Intelligence enables consistent decisions and aligned execution at scale
Outcome of the model
- Adaptive, not rigid, architectures
- Intelligent, not reactive, organizations
- Decisions based on data and context, not intuition
- Scalability without operational complexity
Key message: Enterprise architecture stops being documentation and becomes a strategic advantage for the organization.
References
- Kotusev, S. (2016). Enterprise Architecture Frameworks: The Fad of the Century. British Computer Society (BCS). https://www.bcs.org/articles-opinion-and-research/enterprise-architecture-frameworks-the-fad-of-the-century/ · Author’s PDF copy, pp. 1–5: https://kotusev.com/Enterprise%20Architecture%20Frameworks%20-%20The%20Fad%20of%20the%20Century.pdf