by Daniel Lambert

​

Business and enterprise architects play a pivotal role in shaping how AI is integrated and governed across the enterprise. Their involvement spans all seven layers of the AI model architecture, from infrastructure to end-user applications, as shown in Figure 1 above. While their influence is limited in the more technical, infrastructure-focused layers, it becomes increasingly strategic as AI moves closer to business-facing domains. Architects need to align AI systems with business strategies, ensure interoperability, uphold ethical standards, and support scalability and compliance. By guiding AI development with a focus on value, usability, and trust, they ensure that AI investments drive meaningful outcomes and long-term business success.

​

1. Physical Layer: Hardware & Infrastructure

​

For enterprise architects, involvement at the Physical Layer is generally somewhat limited. Their role primarily involves aligning IT infrastructure strategy with business objectives and ensuring AI workloads can be supported within enterprise architecture frameworks. Architects should maintain a general understanding of infrastructure options, such as cloud platforms (AWS, Azure, GCP), edge computing, and quantum capabilities, to inform higher-level decisions. Collaborating with IT infrastructure teams, they may help define non-functional requirements like scalability, latency, and security, which indirectly influence AI model performance and accessibility.

​

2. Data Link Layer: Data Source & API Integration

​

At the Data Link Layer of an AI model architecture, enterprise architects should ensure that AI model integration aligns with enterprise service architecture and digital transformation strategies. Their responsibility lies in evaluating the model-serving architecture for interoperability, modularity, and maintainability. Enterprise architects will use value streams to identify required information types (or business objects), as shown in Figure 8 of this article. Enterprise architects will also detect the data sources for this information. In addition, architects support API governance, standardization, and the secure exposure of AI capabilities across enterprise systems. They can finally facilitate the alignment of AI service integration with business workflows and IT governance policies, ensuring that AI APIs and pipelines support scalable and resilient business operations.

​

3. Computation Layer: Processing & Logical Execution

​

At the Computation Layer, enterprise architects need to focus on ensuring that the AI execution environments align with enterprise computing strategies. Though they may not design these systems, they assess compatibility between AI models and existing platforms, especially when edge computing or federated learning is involved. Architects collaborate with data engineers and AI teams to ensure appropriate AI frameworks (e.g., TensorFlow, PyTorch) are selected in accordance with enterprise guidelines. Their influence ensures compute resources are deployed where they can best support AI-driven business capabilities.

​

4. Knowledge Layer: Retrieval & Reasoning Engine

​

Enterprise architects play a strategic role in the Knowledge Layer by framing how knowledge assets, external data sources, and internal repositories should be structured and leveraged to enhance AI decision-making. They help establish a governance model around knowledge retrieval technologies, such as RAG systems, knowledge graphs, and vector search tools. By aligning data discovery and reasoning systems with enterprise knowledge management strategies, architects ensure AI capabilities support fact-based decision-making across domains.

​

This layer is particularly critical for enabling business transparency and explainability in AI decisions. Architects are instrumental in integrating AI with enterprise knowledge hubs (e.g., corporate wikis, CRM platforms, legal databases) and ensuring alignment with compliance, data quality, and lifecycle management policies. Their role extends to defining how reasoning engines integrate with domain-specific business logic to create intelligent systems that are trustworthy, context-aware, and scalable. Additionally, they must ensure semantic consistency and business rule alignment across AI-driven retrieval mechanisms, which is vital for enhancing business operations and user trust.

​

5. Learning Layer: Model Training & Optimization

​

The Learning Layer represents a pivotal point where enterprise architects begin to exert significant influence. Although the development of neural networks and optimization algorithms is the domain of data scientists and ML engineers, architects ensure that training initiatives are strategically aligned with enterprise goals and business capabilities. They help define the business contexts and objectives for model training, ensuring training data, algorithms, and performance metrics reflect real-world operational needs.

​

Architects also assist in prioritizing AI use cases based on value streams, profitability, and feasibility. They support the definition of requirements around fairness, transparency, and bias mitigation in AI training processes. Furthermore, they advocate for reusable training components, standardized ML pipelines, and ethical AI frameworks that support sustainable innovation. Architects coordinate with data governance teams to ensure training data is sourced legally, securely, and represents diverse business scenarios. By aligning AI training objectives with enterprise KPIs and regulatory mandates, they enhance the relevance, trust, and long-term usability of the AI models.

​

6. Representation Layer: Data Processing & Feature Engineering

​

At the Representation Layer, enterprise architects become highly involved in ensuring that data preparation and feature engineering are driven by meaningful business context. This layer translates raw data into formats that are usable by AI models, and architects provide the guidance to ensure this transformation aligns with enterprise-wide data standards and business semantics.

​

They contribute by defining metadata standards, ontologies, and taxonomies that help ensure consistency and interoperability across data sources. Architects advocate for feature engineering practices that capture not just technical attributes but also critical business signals, customer behaviors, and operational indicators. Their oversight supports the governance of data enrichment, normalization, and embedding techniques, ensuring they contribute to explainable and ethical AI.

​

Enterprise architects also work to align data processing with enterprise data strategies, including master data management (MDM), data lineage, and data privacy regulations. They collaborate with data engineers and analysts to design pipelines that are auditable, reusable, and aligned with organizational goals. Through their stewardship, the representation layer serves not only technical AI needs but also broader enterprise data integrity and decision-making requirements.

​

7. Application Layer: AI Interface & Deployment

​

The Application Layer is where business and enterprise architects have the greatest influence, as it directly interfaces with business users, customers, and strategic objectives. At this layer, architects play a central role in ensuring AI deployment aligns with digital transformation efforts, customer experience strategies, detailed business capabilities, and operational excellence.

​

They guide how AI is embedded into enterprise applications, customer support systems, chatbots, and business process automation platforms. Architects ensure AI deployments are user-centric, scalable, and compliant with enterprise security and data governance policies. They oversee the integration of AI into workflows, ensuring it improves existing processes and enhances business outcomes.

​

Architects are also responsible for defining KPIs and success metrics for AI applications, working with business leaders to ensure that solutions deliver measurable value. They help manage priorities, change, adoption, and training for users engaging with AI systems. Additionally, they advocate for ethical AI principles and transparent user experiences, particularly in customer-facing solutions. Through governance, strategy alignment, and cross-functional coordination, business and enterprise architects ensure that AI applications deliver long-term business value, support continuous improvement, and maintain trust with stakeholders.

​

 

In summary, the involvement of enterprise architects grows as we move from the technical foundations of AI to its business-facing applications. While their influence is lighter in infrastructure-heavy layers, their strategic, governance, and alignment roles become vital from the knowledge layer upward, ensuring that AI initiatives truly serve enterprise goals and user needs.