Engineering the High-Velocity Path to the Agentic Enterprise

In a recent sit-down, 66degrees CEO Ben Kessler and CTO Brian Gregory analyzed the mechanics of the "Agentic Enterprise." This state is defined by a collaborative ecosystem where human intelligence and autonomous AI agents operate within a unified framework to accelerate organizational growth.

While the vision of the Agentic Enterprise is compelling, the path to execution is frequently obstructed by legacy friction and strategic hesitation. Kessler and Gregory unpacked the specific reasons AI strategies stall, the non-negotiable role of data modernization, and why leadership cannot afford to wait for a "perfect" roadmap to initiate the pilot-to-production lifecycle.

The Execution Gap: Vision vs. Reality

The discussion opened with a reality check regarding the current state of AI adoption. Brian Gregory noted that while approximately 70% of enterprises possess an active AI strategy, only one in ten agentic opportunities successfully transition into a production-grade asset.

This disconnect persists because organizations remain trapped in isolated testing environments. In the early stages of AI, engineering a single-point solution for a narrow problem was considered a success; however, these isolated pilots lack the mechanics required to scale. The true power of the Agentic Enterprise lies in the "multiplier effect" triggered by an agentic framework—allowing autonomous agents to communicate across workflows to solve multi-step business challenges rather than singular tasks.

The Foundation: Data Modernization as a Strategic Asset

A recurring theme throughout the session was the absolute necessity of data liquidity. As Gregory articulated, "You’re only as good as your data." For AI to function as a reliable asset, it requires access to clean, enabled data—yet for many legacy organizations, this "oil" remains trapped in siloed, on-premise systems or mainframes.

Historically, modernizing this data presented a massive operational risk. Migrating legacy code, such as COBOL, was considered high-friction because the original logic was buried in millions of lines of undocumented code. Organizations feared that refactoring these core systems—which govern adjudication, payments, and critical infrastructure—could trigger catastrophic outages.

However, the risk profile has been fundamentally re-engineered. Gregory highlighted how new technical tools, such as Google’s mainframe assessment tools and Gemini, can now analyze legacy code to provide deep insights into logic triggers. This AI-assisted analysis grants IT leaders the rigor to refactor systems, accelerating modernization timelines by 70-80% while mitigating human error.

The Parallel Strategy: Accelerating Velocity

The interview debunked the "waterfall" myth of AI adoption—the belief that data modernization must be 100% complete before AI experimentation begins. Gregory advocates for a parallel execution strategy to maintain competitive velocity:

1. Modernize the Data Estate: Initiate the heavy lift of harmonizing data and refactoring legacy systems to unlock the "oil." Critical data trapped in siloed mainframes must be decoupled to ensure it is "consumable." While migrating COBOL was once high-risk, AI tools now provide the confidence to refactor these systems effectively. You cannot architect an agentic enterprise without clean, accessible data foundations.
2. Launch AI Pilots: Simultaneously execute pilots to solve immediate business friction. Gregory cited a supply chain scenario where an employee manually sources products and awaits cost approvals—a workflow that typically consumes 3 to 4 days. By deploying an automated agent, this process is reduced to mere minutes. These high-ROI pilots do not require a full data transformation; they allow the organization to generate immediate value while the broader infrastructure modernization runs in parallel.

Engineering for Open Standards

Regarding technology partners, the interview highlighted why Google Cloud is the premier platform for the Agentic Enterprise. Beyond the full-stack hardware (TPUs) and reasoning engines (Gemini), the critical differentiator is technical interoperability.

Gregory warned against the long-term cost of closed "walled gardens." In the near future, internal agents must communicate with external agents—such as a restaurant’s inventory agent interfacing directly with a supplier’s fulfillment agent. If these agents are built on proprietary, closed stacks, they cannot interact, neutralizing their strategic value. Google’s focus on open standards ensures the agentic stack remains flexible and future-ready.

Bridging the Gap

While Google provides the tools for scale, a gap often exists between modern capabilities and a client’s legacy constraints. Kessler and Gregory discussed the role of 66degrees in bridging this divide through Paradigm. Set to launch fully in 2026, Paradigm serves as the accelerated framework to help enterprises refactor legacy code while implementing the governance and orchestration required for autonomous agents.

The Bottom Line: Start Yesterday

The technology landscape is shifting quarter-over-quarter; leaders who wait for a "perfect" roadmap will be left flat-footed. The interview concluded with a mandate for urgency: Gregory emphasized that the work of harmonizing data needs to start "yesterday." Ben Kessler reinforced this, advising leaders to reject the trap of perfection: "Don't let perfection stand in the way of progress... Start today." Get in touch and learn how 66degrees transforms Agentic Enterprise complexity into clarity.

‍

‍