Navigating and creating value within Business, Enterprise, Solution and Technical Architectures
by Rick Ross
RMIT University, 2026
Architects are rapidly transitioning from using traditional architecture methods, toward new approaches that incorporate Artificial Intelligence (AI). In an AI enabled economy, increased productivity and efficiency within a firm will directly attribute to further competitive advantage and contribute to a macroeconomic lift in Gross Domestic Product (GDP). The architect is considered as a strategic designer and navigator that provides investment validation at a business and/or technology level. Competent architects develop their craft through a combination of professional experience, industry exposure, and academic learning acquired over a long endured period. The value proposition of a competent architect to a firm is grounded in their ability to curate long-term optionality through informed decision making, future foresight, alignment to business and technology objectives, and ethical governance to ensure accountability. This recipe is Architecture wisdom.
In a world where AI is increasingly used to enhance productivity, the question arises: how will this affect the Architect?
This article explores options for scaling and augmenting the architect’s role, examining benefits and limitations of AI across business, enterprise, and solution architecture continuums.
In this article, architecture continuums refer to layered architectural perspectives that translate organisational intent into operational and technological outcomes. Together, these continuums provide a structured lens for aligning, governing, and operationalising AI across organisations (Ross, Weill, and Robertson 2006; The Open Group 2022).
Business Architecture operates at the highest level of abstraction and focuses on organisational strategy, value creation, and capability design. It establishes the strategic context for AI investment decisions and ensures that AI initiatives reinforce enterprise objectives and stakeholder value rather than isolated optimisation (Ross et al. 2006; Trad 2021).
Enterprise Architecture bridges strategic intent and execution by providing structural coherence across capabilities, processes, information, and governance. It defines shared principles and roadmaps that enable AI adoption to scale consistently while meeting enterprise wide governance and regulatory requirements (Van de Wetering 2022; The Open Group 2022).
Solution Architecture translates enterprise direction into specific system designs that address defined business problems. It governs how AI capabilities are embedded within applications, workflows, and integrations to meet functional and non-functional requirements while supporting controlled system evolution (Dilnutt et al. 2023; Hindarto 2024).
Technical Architecture underpins higher layers by defining the infrastructure, platforms, and technology services required to deploy and operate AI at scale. This includes compute platforms, data architectures, and security foundations that ensure performance, resilience, and compliance across diverse environments (Masuda and Viswanathan 2025; Van de Wetering et al. 2021).
To contextualise the impact of AI on architectural practice, this section contrasts key characteristics of architecture prior to widespread AI adoption with those observed in the current AI augmented era. By contrasting traditional and contemporary architectural practices, the discussion highlights the importance of recognising AI’s long-term societal impacts, balancing human expertise against the potential risk of role displacement. The analysis focuses on key productivity opportunities faced by architects, including faster content generation during options analysis and design, improved stakeholder engagement, and the formulation of current, transition, and target states in environments with variable data quality. In addition, the article examines repeatable architectural activities that can be enhanced through the use of Architecture AI agents.
The Architect’s Evolution: from manual design to AI augmentation across technology eras
The role of the architect has evolved in history. To understand the current shift, consider the architectural evolution described in a series of waves:
Wave 1: The structural era (1960s–1980s): Architecture during the period was physical and centralized. Architects were “Master Builders” of mainframes, focusing on rigid, batch-processed stability through rigid batch-processing environments. Architecture methods were dominated by waterfall system development life cycle (SDLC), requiring complete design before any implementation commenced (Royce 1970; Brooks 1975).
Wave 2: The Functional Era (1990s): The emergence of client‑server computing introduced a focus on interoperability across functional silos. The role of the Solution Architect emerged to bridge these divisions. Architecture frameworks such as Zachman Framework (Zachman 1987). provided structured viewpoints of the enterprise.
Wave 3: The Adaptive Era (2000s–2010s): Cloud computing and Agile delivery models shifted architects from “gatekeepers” to “enablers.” Static blueprints were replaced with API‑driven, modular landscapes. Service‑oriented architecture, microservices, and cloud platforms became foundational, supported by evolved methods such as The Open Group Architecture Framework, TOGAF (The Open Group 2022).
Wave 4: The Generative Era (2025–Present): The current era introduces agentic architecture. Architects no longer solely design systems; they design intelligence loops that enable systems to self‑evolve. Architecture methods are increasingly being rewritten to incorporate AI, enabling near real‑time digital twins of organisations and their technology environments (Patil 2025; Kooy, Piest, and Bemthuis 2025; Ardoq 2024; DiVA Portal 2024).
Across all eras, the absence of effective architecture introduces significant cost, financial, societal, and technical debt among others. This often results in “accidental architecture.” This manifests through reactive and uncoordinated outcomes rather than deliberate, strategic planning, commonly driven by tactical decision‑making, insufficient integration oversight, or inconsistent architectural engagement.
AI adoption barriers for the Architect: Global context
Global research and industry studies consistently indicate that the primary barriers to successful AI adoption are organisational and architectural rather than technological. Across both mature and emerging economies, a majority of organisations report experimenting with artificial intelligence; however, only a minority achieve sustained, enterprise‑wide productivity gains (OECD, BCG, & INSEAD, 2025; Li, Zhu, & Hua, 2025). This divergence between adoption and realised value suggests that AI initiatives implemented without coherent enterprise architectural integration often fail to scale beyond isolated use cases.
One commonly observed barrier is regulatory ambiguity, as fragmented and evolving AI governance regimes across jurisdictions create uncertainty for executive investment decisions and complicate compliance for multinational organisations (Kremer et al., 2023; OECD et al., 2025). In parallel, organisations frequently encounter the “J‑Curve” of intangible capital, where short‑term productivity declines associated with workforce reskilling, process redesign, and legacy system modernisation suppress early returns and dampen momentum (OECD et al., 2025).
A further global constraint is the human‑capital gap, particularly the shortage of professionals capable of translating opaque AI outputs into explainable, governed, and contextually valid architectural decisions. Empirical evidence indicates that without enterprise architecture capability spanning organisational context, ethics, and governance, AI adoption remains fragmented and vulnerable to risk, bias, and value leakage (Li et al., 2025; OECD et al., 2025).
AI benefits for the Architect: Global context
The benefits of AI adoption for architects are increasingly framed around translating experimentation into measurable productivity gains. While a majority of organisations globally report having experimented with AI, only a minority have achieved enterprise wide productivity uplift, largely due to fragmented architecture, weak data foundations, and governance constraints (OECD, BCG, and INSEAD 2025; Li, Zhu, and Hua 2025). Within this environment, architects play a critical value creation role by aligning AI investments to business strategy, enterprise capability models, and regulated operating environments.
At the business architecture layer, AI enabled analytics support improved investment validation and strategic decision making, addressing executive concerns around return on investment and risk transparency. At the enterprise architecture level, architects leverage AI to establish a coherent AI roadmap, capability alignment, and responsible AI guardrails, an increasingly important function given the global emphasis on ethical governance, explainability, and regulatory accountability across jurisdictions (Kremer et al. 2023; World Economic Forum 2024).
At the solution and technical layers, AI adoption enables process optimisation, cost reduction, and scalable digital platforms. However, these benefits are only realised when architects actively curate, govern, and contextualise AI outputs to reflect organisational reality rather than theoretical optimisation. Collectively, this positions the Architect not merely as a system designer, but as a strategic integrator who converts AI experimentation into sustained, compliant, and economically meaningful productivity outcomes. A summary of these benefits in the use of AI is included within Table 1.
Table 1 AI benefits for the Architect Matrix
| Architecture layers | Role in AI Integration | Key Benefits | Source |
|---|---|---|---|
| Business | Aligns AI initiatives with strategic objectives and business value creation. | – Enhanced decision-making and innovation alignment. | Trad (2021); Olaniyi (2024) |
| Enterprise | Aligns AI initiatives with enterprise strategy and business capabilities | – Consistent AI vision and roadmap – Responsible and transparent AI operations | Van de Wetering et al. (2021); Trad (2021) |
| Solution | Streamlines processes through AI-driven optimization | – Improved productivity and reduced costs – Seamless automation and cross-functional system collaboration | Dilnutt et al. (2023); Hindarto (2024) |
| Technical | Provides infrastructure for AI deployment, including cloud, IoT, and analytics platforms | – Scalable, secure, and adaptive AI-enabled systems. | Van de Wetering et al. (2021); Masuda & Viswanathan (2025) |
AI limitations for the Architect: Global context
Despite the productivity and scale advantages introduced by AI, its application within architectural practice is subject to important limitations that constrain automation and reinforce the necessity of human oversight. Globally, AI systems demonstrate limited ability to internalise organisational context, institutional history, and political dynamics that frequently shape architectural decision making. While AI can optimise against explicit constraints such as cost or performance, it struggles to reason about implicit factors including stakeholder incentives, cultural resistance, and transitional feasibility. As a result, AI generated architectural recommendations may be theoretically optimal yet misaligned with organisational reality (Li, Zhu, and Hua 2025).
AI effectiveness is further constrained by dependency on data quality, completeness, and provenance. Fragmented enterprise repositories and inconsistent metadata can cause AI systems to propagate inaccuracies at scale, increasing the risk of architectural drift when outputs are insufficiently validated (OECD, BCG, and INSEAD 2025).
Additional limitations arise in explainability and accountability. Architectural decisions require traceability, defensibility, and liability ownership, particularly in regulated environments, responsibilities that remain inherently human despite advances in responsible AI (Kremer et al. 2023; World Economic Forum 2024). Furthermore, automation of foundational architectural activities may weaken capability development for early career architects if roles and learning pathways are not deliberately redesigned. Together, these limitations reinforce that AI augments architectural capability but does not replace architectural responsibility. A summary of these limitations is provided in Table 2.
Table 2 AI limitations for the Architect Matrix
| Architecture perspective | Limitation | Implication for the architect | Source |
|---|---|---|---|
| Contextual reasoning | Limited understanding of organisational politics and culture | Requires human interpretation and decision arbitration | Li, Zhu, and Hua (2025) |
| Data dependency | Sensitivity to incomplete or low quality enterprise data | Necessitates architect led validation and correction | OECD, BCG, and INSEAD (2025) |
| Governance and accountability | Limited explainability and inability to assume liability | Architect remains accountable for compliance and ethics | Kremer et al. (2023); World Economic Forum (2024) |
| Capability development | Automation of entry level architectural tasks | Risk of long term erosion of architectural expertise | OECD et al. (2025) |
Architecture productivity opportunities using AI
Recent academic and industry research increasingly positions artificial intelligence as a general purpose capability that reshapes how architectural work is performed rather than simply automating isolated tasks. In the context of enterprise architecture, AI is shown to materially improve productivity by expanding the range of design options explored, accelerating architectural sense making, and reducing the manual effort associated with maintaining architectural artefacts (Kooy, Piest, and Bemthuis 2025). These effects are particularly visible in environments characterised by architectural complexity, legacy inertia, and high rates of change. Figure 1 illustrates how AI increases architectural throughput while preserving human responsibility as the governing constraint between analysis and decision making.
Source: Author, 2026
Generative AI techniques have been empirically demonstrated to support architectural design ideation and options analysis by enabling architects to explore multiple feasible design alternatives within defined constraints. Rather than replacing architectural judgment, generative approaches increase the breadth of alternatives considered and improve trade off awareness across cost, risk, and performance dimensions (Peckham et al. 2025; Kooy et al. 2025). This justifies the shift of the architect’s role from sole author toward curator and decision arbiter, as described in this article.
Similarly, research shows that AI mediated translation of technical architecture information into stakeholder specific representations improves decision quality and engagement. Studies in enterprise architecture practice highlight that executive resistance to architecture initiatives often stems from misaligned communication rather than technical validity. AI enabled summarisation, persona based visualisation, and narrative generation reduce this translation burden and reposition architects as strategic partners in decision making (Vanrechem 2025; Kooy et al. 2025).
The use of AI for data quality assessment and architectural state modelling is also supported by emerging evidence. Digital twin and enterprise architecture research demonstrates that AI assisted discovery can identify inconsistencies across fragmented repositories and support more accurate modelling of current, transition, and target states. This capability is critical in large organisations where architectural knowledge is distributed across tools, teams, and undocumented operational practices (Edrisi et al. 2024). Without such capabilities, architecture efforts remain dependent on static documentation that degrades rapidly under change.
Finally, the deployment of AI agents within architectural workflows is increasingly recognised as a means of automating low creativity, high repetition activities such as compliance mapping, repository maintenance, and drift detection. Agentic AI research shows that goal directed agents operating under human oversight can maintain architectural artefacts continuously, enabling architects to focus on higher value activities such as negotiation, governance, and ethical stewardship (Natarajan and Ponnusamy 2025; Bandara et al. 2025). This supports the claim that AI acts as an enabling substrate rather than a replacement for architectural responsibility.
1. Faster content generation and options analysis
AI enables generative design approaches in which architects define constraints such as budget, risk, and latency, allowing AI systems to generate numerous feasible architectural patterns. The architect’s role shifts from author to curator, ensuring outputs are validated against organisational, political, and operational realities to avoid architectural erosion.
2. Improved stakeholder engagement
AI can function as a translation layer, transforming complex technical information into persona‑based views tailored to executive stakeholders. This enhances architecture’s visibility as a strategic partner rather than a back‑office function.
3. Identifying data quality and transition state formulation
Architectural decision making is dependent on data quality. AI can act as a forensic tool, identifying inconsistencies across legacy repositories and supporting more accurate modelling of current, transition, and target states.
4. AI Agent(s) to enhance repeatable work and workflows
Architecture AI agents can automate low‑creativity, high‑repetition tasks such as compliance mapping, metadata tagging, and real‑time drift detection. This enables architects to focus on higher value activities, including strategic negotiation and ethical stewardship.
Augmenting the Architect: The convergence of humans and technology
The future evolution of the architect lies in the convergence of human expertise with AI enabled technologies. Generative AI supports rapid content creation, while AI agents provide autonomous environmental discovery under human oversight illustrated in Figure 2.
Source: Patil, 2025
In this hybrid model, AI agents act as the “intelligent fabric” or “telemetry probes” of the enterprise, reaching into the physical implementation layer scanning cloud environments and network configurations to discover the actualized current state in real-time. This eliminates the reliance on outdated documentation, forming an “existing baseline” that is mathematically accurate.
Once this baseline is established, Generative AI models synthesize this raw data into architectural artifacts. However, this automation is specifically designed to provide the cognitive space for the architect to be curiously analytical. With the routine “discovery” work completed by agents, the human architect focuses on high-order “What-If” analysis. By utilizing a modern Architecture Repository, architects can capture these various transition states, allowing them to compare multiple scenarios side-by-side and validate AI hypotheses against human business intuition.
The Architecture Repository as a Digital Twin
The modern Architecture Repository has evolved beyond a static documentation store to function as a core enabling capability of the Digital Twin of the Organization (DTO). A contemporary repository acts as an authoritative source of truth that maps organizational concepts such as capabilities, processes, applications, and technologies into structured and interconnected software representations. Architecture‑focused repository platforms including Ardoq, SAP LeanIX, Orbus Software (iServer / OrbusInfinity), Avolution ABACUS, and Sparx Systems Enterprise Architect amongst others explicitly support this role by maintaining formal metamodels, dependency structures, and continuously curated architectural data that enable analysis and architectural reasoning over time (Ardoq 2024; SAP LeanIX 2024; Orbus Software 2024; Avolution 2025; Sparx Systems 2025; DiVA Portal 2024).
In parallel, a growing number of enterprise IT and operational platform vendors are extending historically operational systems into enterprise architecture and digital twin domains. Platforms such as ServiceNow, BMC Helix, and Atlassian Jira Service Management have introduced enterprise architecture and service modelling capabilities built on configuration management databases and common service data models. These platforms emphasise real‑time operational visibility, service dependency mapping, and lifecycle governance, enabling forms of organisational digital twins when supplemented with architectural abstractions and governance frameworks rather than architectural intent alone (ServiceNow 2025; Quintica 2024).
Source: Author, 2026
By creating a dynamic software model that mirrors the organization’s physical and logical assets, the architecture repository establishes a continuous feedback loop between strategy and execution. When paired with automated discovery mechanisms or real‑time data from AI agents, the repository increasingly exhibits the characteristics of a living digital twin, enabling architects to anticipate bottlenecks, identify redundant or duplicate systems, and assess transformation impacts before they materially affect operational performance or economic outcomes (Edrisi et al. 2024; Ardoq 2024). Figure 3 illustrates the conceptual digital twin components for all architectural layers.
Societal Impact and Conclusion
As productivity increases, the risk of displacement for junior roles is a reality that must be acknowledged, particularly as automation absorbs entry‑level analytical and documentation tasks (OECD, BCG, & INSEAD 2025; Li, Zhu, & Hua 2025). However, the “Ascension” of the architect suggests that while AI can increasingly design the “Perfect State,” it cannot yet negotiate the “Possible State” within complex, human‑oriented organisations shaped by politics, ethics, and social legitimacy (Li et al. 2025; World Economic Forum 2024). The future therefore belongs to the Augmented Architect who balances technological speed with ethical transparency, accountable decision ownership, and human agency (Kremer et al. 2023; World Economic Forum 2024).
References
Ardoq. 2024. “Evolution of EA: Digital Twin of an Organization.” Ardoq Blog, November 25, 2024. https://www.ardoq.com/blog/digital-twin-of-an-organization.
Atlassian. (2024). ITSM use cases and reference architectures. Atlassian.
Avolution. (2025). Digital twin architecture with ABACUS. Avolution Software. https://www.avolutionsoftware.com/digital-twin/
Bandara, E., Gorea, R., Foytik, P., Shetty, S., Mukkamala, R., Rahman, A., Liang, X., Bouka, S., Hass, A., and Rajapakse, S. (2025). A practical guide for designing, developing, and deploying production grade agentic AI workflows. arXiv. https://arxiv.org/abs/2512.08769
BMC. (2025). BMC Helix ITSM architecture overview. BMC Software.
Brooks, F. P. (1975). The mythical man-month: Essays on software engineering. Addison-Wesley.
CEDA (Committee for Economic Development of Australia). 2025. “Smaller Australian Businesses Are Missing Out on AI. It’s Time to Fix That.” Opinion, November 26, 2025. https://www.ceda.com.au/news-and-resources/opinion/technology/smaller-australian-businesses-are-missing-out-on-ai-its-time-to-fix-that.
Dilnutt, R., Xiao, D., Fang, G., Yang, J., Ran, R., & Wu, L. 2023. How Does Enterprise Architecture Support Artificial Intelligence-Involved Digital Transformation in The Manufacturing Industry.
DiVA Portal. 2024. “Developing and Evolving a Digital Twin of the Organization.” Volume 12, 2024. https://www.diva-portal.org/smash/get/diva2:1853204/FULLTEXT01.pdf.
Edrisi, F., Perez Palacin, D., Caporuscio, M., and Giussani, S. (2024). Developing and evolving a digital twin of the organization. IEEE Access, 12, 45475 to 45494. https://doi.org/10.1109/ACCESS.2024.3381778
Governance Institute of Australia. 2025. 2025 AI Deployment and Governance Survey Report. April 2025. https://www.governanceinstitute.com.au/app/uploads/2025/04/AI-deployment-and-governance.pdf.
Hindarto, D. 2024. Building the future of the apparel industry: The digital revolution in enterprise architecture. Sinkron: jurnal dan penelitian teknik informatika, 8(1), 542-555
Kooy, Stefan Julian, Jean Paul Sebastian Piest, and Rob Henk Bemthuis. 2025. “Impact and Implications of Generative AI for Enterprise Architects in Agile Environments: A Systematic Literature Review.” arXiv preprint arXiv:2510.22003. https://doi.org/10.48550/arXiv.2510.22003.
Kremer, A., Luget, A., Mikkelsen, D., Soller, H., Strandell‑Jansson, M., & Zingg, S. (2023). Governance and regulation as generative AI advances. McKinsey & Company. https://www.mckinsey.com/capabilities/risk-and-resilience/our-insights/as-gen-ai-advances-regulators-and-risk-functions-rush-to-keep-pace
Li, J., Zhu, F., & Hua, P. (2025). Overcoming the organizational barriers to AI adoption. Harvard Business Review. https://hbr.org/2025/11/overcoming-the-organizational-barriers-to-ai-adoption
Masuda, Y., & Viswanathan, M. 2025. Direction of digital it and enterprise architecture. In Enterprise Architecture for Global Companies in a Digital IT Era: Adaptive Integrated Digital Architecture Framework (AIDAF) (pp. 21-70). Singapore: Springer Nature Singapore.
Natarajan, S., and Ponnusamy, A. (2025). Agentic AI architecture framework for enterprises. InfoQ. https://www.infoq.com/articles/agentic-ai-architecture-framework
OECD, BCG, & INSEAD. (2025). The adoption of artificial intelligence in firms: New evidence for policymaking. OECD Publishing. https://doi.org/10.1787/f9ef33c3-en
Olaniyi, G. (2024). Enterprise Architects Leveraging AI for Business Innovation. Available at SSRN 5033526
Patil SP. Leading Through Innovation: The Evolving Role of the Architect in the Cloud and AI Era. J Artif Intell Mach Learn & Data Sci 2025 3(1), 3129-3133. doi.org/10.51219/JAIMLD/sandeep-parshuram-patil/640
Quintica. (2024). Deep dive into ServiceNow’s enterprise architecture capabilities. Quintica.
Peckham, O., Raines, J., Bulsink, E., Goudswaard, M., Gopsill, J., Barton, D., Nassehi, A., and Hicks, B. (2025). Artificial intelligence in generative design A structured review of trends and opportunities. Designs, 9(4), 79. https://doi.org/10.3390/designs9040079
RBA (Reserve Bank of Australia). 2025. “Technology Investment and AI: What Are Firms Telling Us?” Bulletin, November 13, 2025. https://www.rba.gov.au/publications/bulletin/2025/nov/technology-investment-and-ai-what-are-firms-telling-us.html.
Ross, J. W., Weill, P., and Robertson, D. C. (2006). Enterprise architecture as strategy Creating a foundation for business execution. Harvard Business School Press.
Royce, W. W. (1970). Managing the development of large software systems. Proceedings of the IEEE Western Electronic Show and Convention (WESCON), 1–9.
ServiceNow. (2025). Enterprise architecture on the Now Platform. ServiceNow.
https://www.servicenow.com/products/enterprise-architecture.htm
Sparx Systems. (2025). Architecture repository in Enterprise Architect. Sparx Systems. https://sparxsystems.com/enterprise_architect_user_guide/17.1/guide_books/ea_architecture_repository.html
The Open Group. (2022). TOGAF Standard Version 10. The Open Group. https://www.opengroup.org/togaf
Trad, A. 2021. The business transformation enterprise architecture framework for innovation: The role of artificial intelligence in the global business education (RAIGBE). The Business & Management Review, 12(1), 8 2 – 9 7.
Van de Wetering, R. 2022. The role of enterprise architecture-driven dynamic capabilities and operational digital ambidexterity in driving business value under the COVID-19 shock. Heliyon, 8(11).
Vanrechem, S. (2025). How agentic AI elevates the enterprise architect’s role. Architecture and Governance Magazine. https://www.architectureandgovernance.com/applications-technology/how-agentic-ai-elevates-the-enterprise-architects-role
Zachman, J. A. (1987). A framework for information systems architecture. IBM Systems Journal, 26(3), 276–292.
Copyright: © 2026 Rick Ross.
