Building Modern Enterprise Architectures

Enterprise Architecture in an Uncertain World

We see significant volatility on the horizon across geopolitical and macroeconomic fronts, raising critical questions about how these forces shape enterprise architecture decisions and long-term technology planning.

Two opposing but equally crucial forces are now at play:

1. Flexibility and agility in architecture to rapidly adapt to new risks and opportunities
2. Leverage and cost efficiency in IT operations to maintain financial discipline

Balancing these forces requires us to navigate complexity, rethink resilience, and position technology as a strategic enabler of competitive advantage.

This discussion will explore how enterprise architecture must evolve to remain resilient, cost-effective, and AI-ready—while ensuring long-term compliance, security, and adaptability.

Key Discussion Areas

1. Geopolitical & Macroeconomic Exposure: Identifying Risks and Opportunities

• What macro shifts—supply chain disruptions, deglobalization, protectionist trade policies—are creating risks for IT strategy?
• How should enterprise architecture adapt to a more fragmented global technology landscape (e.g., regional cloud strategies, sovereign data requirements, diversified supply chains)?
• Where are companies seeing opportunities for competitive advantage despite uncertainty?

2. Compliance & Long-Term Regulatory Durability

• With AI and data regulations evolving rapidly (e.g., AI Act in Europe, China’s data sovereignty laws, US cybersecurity mandates), how are organizations designing architectures that ensure regulatory resilience?
• How are we baselining compliance thresholds for cross-border data governance, cybersecurity frameworks, and AI ethics?
• What lessons can be learned from industries already deeply regulated (e.g., financial services, healthcare) in embedding compliance into enterprise architectures?

3. Rethinking Operational Exposure & Supply Chain Resilience

• As supply chains become more digital and interdependent, how are we reassessing vendor risk, cloud exposure, and software supply chain security?
• Are organizations shifting toward multi-cloud and hybrid strategies to mitigate concentration risk?
• How are companies embedding business continuity and disaster recovery (BCDR) frameworks into IT architectures to hedge against geopolitical risks?

4. Capital Allocation: Balancing Stability with AI-Driven Innovation

• How are we reallocating budgets to address both resilience (stability) and AI transformation (growth)?
• What changes are being made to IT investment strategies to support the increasing costs of AI infrastructure and compute power?
• How are companies making build vs. buy vs. source decisions for AI and automation?

5. Architecting for AI: The Core vs. Edge Tradeoff

• As AI capabilities mature, where should intelligence sit in the architecture—centralized at the core or distributed at the edge?
• What trade-offs exist between a thin core (highly distributed AI) vs. a thick core (centralized intelligence with AI governance)?
• How does this decision impact latency, security, and scalability?

6. Enterprise Architecture in the Age of AI: Build vs. Buy vs. Source

• What should be built internally for differentiation vs. leveraged externally for capital efficiency?
• How are we determining where to invest in proprietary technology vs. using off-the-shelf AI models and SaaS solutions?
• What lessons have been learned in striking the right balance between agility, cost, and control?

7. Cybersecurity & National Security: The Next Frontier of Governance

• How are enterprise architectures evolving to meet new cybersecurity challenges in an era of AI-driven threats?
• What strategies are emerging for zero-trust security, AI model integrity, and AI-driven cyber defense?
• How are organizations designing architecture frameworks that align with national security regulations and evolving global cyber policies?

GE Vernova Case Study: A ~$35B Public “Startup” Rebuilding Enterprise Architecture from Scratch

As a recently spun-off entity from GE, GE Vernova had the unique opportunity to design its enterprise architecture without legacy constraints. This session will examine:

• Key decisions in defining a modern enterprise architecture
• Trade-offs and roadblocks encountered
• Lessons learned in building for agility, AI-readiness, and cost efficiency

Following the case study, other members will share perspectives on their own enterprise architecture challenges, exploring themes such as:

• AI & data architecture evolution
• Modern SaaS strategy and cloud optimization
• Enterprise governance in volatile environments
• Business transformation through AI and automation

Architecting for Autonomous Processing and Operating Models

Getting Autonomous Processing and Operating Models right

A blog on (one view of) where Agentic AI eventually lands follows below. The journey from “here” to (some version of) “there” will clearly require broad-based changes across technology infrastructure, data foundations, and AI deployments. But just as critically, it demands shifts in AI adoption, change management, process redesign, operating model transformations, and strong cross-enterprise leadership.

As organizations move beyond proofs of concept and isolated use cases, CIOs must confront key questions:

• How do we ensure AI investments yield real business value and scale sustainably?
• How do we integrate AI into mission-critical workflows without disrupting current operations?
• How do we evolve governance and control frameworks to balance innovation and risk?
• How do we manage talent, leadership, and change effectively to embed AI into how the enterprise runs?

This session will explore where we are today, what’s working, and what must improve as we advance toward AI-driven operating models and autonomous enterprise architectures.

Key Discussion Areas

1. Assessing Where We Are in the Journey: Progress and Challenges

• What AI initiatives have demonstrated tangible business impact today?
• Where are organizations still stuck—data fragmentation, model accuracy, lack of business alignment?
• What industries or functions have seen the most success so far in moving from pilots to scaled AI adoption?

2. Scaling AI: Moving from Pilots to Full-Scale Deployments

• What organizational barriers (e.g., process inertia, workforce skills, regulatory concerns) hinder scaling AI beyond proof of concept?
• How do we measure and communicate ROI to justify continued investment?
• How do we design AI-driven architectures that are scalable, cost-effective, and maintainable?

3. Lessons from Deploying Generative AI at Scale

• What has worked (and not worked) in enterprise-wide AI adoption?
• How do we manage the gap between AI capabilities and user trust/adoption?
• What process changes, automation frameworks, and AI-specific operating model shifts have helped drive adoption?

4. AI Guardrails and Governance: Controlling the Next Phase of AI Evolution

• What governance models are emerging to balance AI-driven autonomy with control, compliance, and ethics?
• How are organizations ensuring data privacy, security, and regulatory adherence in AI-driven workflows?
• What’s the role of centralized AI governance vs. distributed ownership across business units?

5. Balancing Today’s Operating Model with Designing Tomorrow’s Enterprise

• How do we allocate capital, resources, and leadership focus between running today’s business and designing for the AI-first future?
• How do we redesign enterprise processes for agentic AI without disrupting current operations?
• What strategies work to build an AI-ready workforce, leadership mindset, and organizational agility?

Why the Buzz on Agentic AI

(a LinkedIn Blog)

There is a buzz around Agentic AI that's worth unpacking:

What is Agentic AI?

Its best to start with an example — in planning for my trip last week, here are three questions I could have asked:

1. “What’s the weather in Jaipur?”
2. “Should I bring an overcoat to Jaipur for this week?”
3. “Can you put together my packing list for Jaipur?”

Each question represents a step in the evolution of AI capabilities:

• The first question leverages machine learning and NLP (Natural Language Processing). It pulls structured data (weather forecasts) and uses NLP to interpret your query. This is AI at its most familiar and has been around for a while—fetching information and presenting it back to you.
• The second question adds nuance. “Should I bring an overcoat?” requires Generative AI—which interprets not just the data (Jaipur’s weather) but the context (day vs. night temperatures and maybe even knowledge of fashion trends). It generates an informed suggestion rather than merely reporting facts - and we have been using this since ChatGPT first came out.
• The third question shifts into Agentic AI—a fundamentally different paradigm. Here’s what happens: It retrieves the weather forecast and checks your calendar (business or leisure?). It examines your travel history to determine if you prefer carry-on or check-in luggage. It builds an initial packing list, but the process doesn’t stop there. As the agent assembles your list, it realizes the items won’t fit into your carry-on. So, it iterates—removing and replacing items, rechecking for weight and volume, until the list works. Noticing you’re missing an evening tie, it spins up another agent to log into your account at Nordstrom, find the right tie, and order it to arrive before your departure.

This is Agentic AI in action: not just helping you make decisions but making them for you—autonomously, contextually, and with adaptability.

The rise of Agentic AI represents a monumental leap forward— what’s becoming really obvious:

Technology Gets Exponentially Better

• Scale will be Massive. We’re entering the era of the Multi-Agentic Workforce, where AI agents will vastly outnumber human employees. During a recent discussion, NVIDIA’s CEO predicted, “We will have 50,000 employees and 100 million agents in two years.” This isn’t a 5x or 10x improvement over previous automation tools like RPA—this is 100x the scale. Every repetitive task, operational bottleneck, and even complex decision-making processes can be handed off to agents at unprecedented volumes.
• Performance becomes Transformational. Agentic AI doesn’t just execute tasks faster; it thinks smarter. These agents will leverage “long thinking” to solve complex, multi-step problems. For example, instead of simply automating a report, an agent can analyze trends over months, predict risks, and propose strategic responses. They will also feature dynamic adaptability, pulling humans into the loop for decisions that require creativity or moral judgment, while handling the rest autonomously.
• Organizations will become Self-Driving. Imagine a business where agents don’t just follow orders—they actively manage themselves. Agentic AI will allow for agents to autonomously spawn, train, deploy, monitor, and retire each other based on organizational needs. They will form self-sustaining ecosystems, reacting to new demands, scaling up for opportunities, or scaling back when priorities shift—all without requiring human micromanagement.

Agentic AI won’t just improve productivity—it will completely rewrite the rules for how organizations are structured and how work gets done.

Future of Work Radically Changes

• Traditional Silos will Blur. Departments like sales, service, and operations have long existed as separate entities, each with its own processes and teams. In an agentic workforce, these silos will dissolve. Agents don’t see “departments”—they see end to end tasks. As a banking COO in Singapore recently noted: “Sales and service are two departments for us today but one for an agentic workforce.” Organizations will evolve from rigid hierarchies into fluid, interconnected ecosystems where tasks flow seamlessly between agents and humans.
• Arc of Talent will Evolve. As agents take over execution and iteration, human talent will shift toward creativity, strategy, and decision-making. As a consumer goods CIO and GBS leader in London noted: "Leaders will need to develop hybrid skills that blend technical fluency with business expertise”. IT departments will become HR department for agents, managing the lifecycle of digital workers—from onboarding to identity and entitlements to performance management. The most successful organizations will foster a culture of partnership between human ingenuity and agentic efficiency.
• Management will go from Command to Coordination. The traditional “command and control” model of management will give way to a focus on coordination and collaboration. Instead of micromanaging tasks, leaders will design workflows that integrate agents into teams and foster collaboration between humans and AI. At the Executive Technology Board, we discussed: “Success will depend on a new set of management skills” - adaptability, ethical decision-making, and the ability to strategically align agentic capabilities with organizational goals.

So what's the Big Deal?

In the technology industry, for the first time, AI agents are breaking out of the traditional software category. While the global Software-as-a-Service (SaaS) market is projected to reach $300 billion by 2030, the global labor market is worth over $50 trillion. Agentic AI shifts the conversation from software as a tool to AI as digital labor, capable of performing autonomous, iterative tasks that were once the sole domain of humans.

From the point of view of the traditional large enterprise, this isn’t just a step forward for automation—it’s a complete redefinition of how businesses allocate resources, and work gets done. Companies won’t just deploy tools to assist their employees; they’ll employ agents capable of handling complex tasks, scaling labor, and adapting in real-time to organizational needs. We are seeing the start of what will become one of the largest economic revolutions in history—the age of digital labor.

Executive Technology Board (c)