Navigating the AI Revolution: A New Blueprint for Business Architecture

A Profession that Must Adapt

I’m a business architect and my profession is never going to be the same again.

Traditionally, the business architect serves as the key player in translating strategic vision into actionable and achievable business models and processes. However, the rise of artificial intelligence (AI), is compelling business architects to explore a new paradigm where AI is not just a tool but a fundamental component of strategic planning and execution.

The last two years have seen game changing leaps in the capabilities of AI, especially Generative AI as seen commercialized in large language models (LLMs) such as OpenAI’s GPT-4. Moreover, the parallel development of complementary tools allows Generative AI models to be economically integrated with business processes and decision-making frameworks. The potential is huge; In a recent study the IMF found that in developed countries, about 60 percent of jobs are expected to be affected by AI.

This presents unparalleled opportunities and formidable challenges to organizations and business architects.  For instance, AI techniques can be used emulate expert-level knowledge management abilities to significantly improve information availability and flows across the organization. Similarly, commodity AI models’ ability to understand and generalize business contexts make available up automation opportunities previously unobtainable due to cost and effort.

Yet, the business architect also needs to be aware of the limitations of AI, where deep rabbit holes await to ensnare the unwary.

A Wealth of Possibilities

The integration of AI technologies into business architecture is prompting a shift from traditional, often siloed, planning and development approaches to more integrated, dynamic, and responsive models. Consider the example of a government department leveraging AI to plan and prioritize its project portfolio. AI can help portfolio planners at all levels and stages, including assisting senior management with ideation and the establishment of priorities, and helping managers perform such duties as due diligence through predictive analytics and the analysis of lessons learned.  Indeed, at the enterprise level, AI can coordinate planning inputs from across multiple organizational service lines to help create a more resilient and achievable portfolio.

With its ability to integrate large amounts of data from different sources AI gives the business architect unprecedented opportunity to design processes that are not dependent on traditional exclusively human centric execution. It’s now possible to create AI agents and imbue them with “personas” to give them a specific business purpose. For example, it’s possible to create a data analyst agent or a report writer agent that are responsible for tasks related to these roles. To execute on their roles, agents are given access to “skills” that they can autonomously call upon to help perform their role. Just as in an actual business organization, agents (such as analyst and report writer) can communicate and exchange draft work products with each other. In their dialogue, agents can iteratively improve the work product or incorporate new information as needed until the task is complete. Orchestration agents can be responsible for coordinating multiple agent-assigned tasks while user proxy agents would be responsible for interfacing with a human supervisor for critical decisions.

From business architect’s perspective, AI represents an additional abstractive cognitive layer that can be incorporated into enterprise processes in two ways. The first is a helper mode.  Here, AI is inserted into a traditional business process as an assistive component to humans. This is akin to having an expert helper on hand for to speed up one or more parts of a human-oriented task. The human would typically explicitly invoke the AI to contribute to a knowledge processing or analytical task. The AI takes the human’s input in the form of data and a prompt and creates the desired output that the human then uses for the next step in the process.

Alternatively, AI can be functionally oriented. In this mode, AI can act across major parts of a business process or function or even its entire length. What would have traditionally involved multiple human actors in different roles can be instead performed by an orchestrated team of specialized agents. The human role would be to validate the output and to ensure that the AI was acting within its guardrails. To be sure, this mode of integration is still in the early stages of development and experimentation and few hard examples of businesses invoking this mode can be found. Yet, the business demand to realize this capability is high so mainstream availability is unlikely to be very far downstream.

Straddling the Jagged Frontier: Grounded Realism is Still a Must

While the integration of AI and business may be a futurist’s dream, the business architect must retain a grounded perspective. Above all, any proposed AI-enhanced business architecture must ultimately be achievable and practical.

Business architects need to be aware that AI is not an efficiency panacea (at least not yet). AI is fraught with limitations, and it can be sometimes difficult to easily know what is straightforward to achieve versus what would be highly complex and expensive.

In his excellent blog article: Centaurs and Cyborgs on the Jagged Frontier, Wharton professor Ethan Mollick explores how tasks that are of equal difficulty from a human perspective may have greatly differing levels of achievability for AI. The business architect needs to understand the principles that define where a task lies on this “jagged frontier” of achievability and whether leveraging AI is a practical approach for that task.

Another key consideration is the potential cultural resistance to AI adoption. For example, employees may fear job displacement or mistrust AI-driven decisions. Business architects must navigate these concerns through inclusive change management strategies that emphasize upskilling, transparency, and the complementary role of AI in augmenting human capabilities.

Other factors sensitive to AI integration include resourcing models, governance and compliance, privacy, and security, and even union considerations. All topics for future articles.

Wrapping Up

In the age of AI, productivity takes on new dimensions. It's no longer just about doing things faster or cheaper; it's about doing them smarter and more strategically. Business architects play a crucial role in redefining productivity metrics to capture the value added by AI, such as improved customer satisfaction, and greater productivity and performance.

As we navigate the age of AI, the role of the business architect is undergoing a profound transformation. No longer confined to traditional models of business process and strategy design, business architects are now at the forefront of integrating AI into the very DNA of organizational strategy and operations. This requires not only a deep understanding of AI technologies and their potential impacts but also a nuanced appreciation of the socio-technical dynamics at play.

The future of business architecture in the age of AI is both promising and challenging. It calls for Business Architects who are not only technologically savvy but also visionary in aligning AI capabilities with strategic objectives, ethical in their approach to data and privacy, and empathetic in managing the human dimensions of AI integration.

Originally published on LinkedIn, February 5, 2024.