Agentic AI Revolutionizes Businesses: A Colossal Challenge

The Rise of AI Agents and the Organizational Challenge

The rapid adoption of artificial intelligence (AI) agents in the business world is creating a notable gap between companies' ambitions and their ability to achieve them. While 85% of organizations have expressed their intention to become agentic within the next three years, 76% of them acknowledge that their current infrastructures are not ready for such a change. This lack of preparedness manifests through gaps in human skills, processes, and workflows, posing a major obstacle to the effective integration of AI.

The Patchwork Problem

The primary challenge companies face lies in their tendency to layer AI agents onto existing structures without thoroughly rethinking their operational model. Prasun Shah, Global Chief Technology Officer for Workforce Consulting and Head of AI at PwC UK Consulting, explains that this approach amounts to integrating AI agents into systems designed for human operations. He compares it to adding tape to a declining operational model, which prevents organizations from fully leveraging the potential of agentic AI.

This method risks stifling the full exploitation of agentic AI's potential, which relies on the ability of agents to manage complex workflows with minimal human intervention. These agents are capable of coordinating tasks, making autonomous decisions, and adapting to changing conditions, which could accelerate business processes by 30% to 50% and reduce time spent on low-value tasks by 25% to 40%.

In experimental areas such as customer service, human resources, and sales, the potential impact of AI agents is already visible. However, for this capability to be fully realized, an enterprise-wide change is necessary, which involves a revision of current organizational structures.

Enriching the AI Vocabulary

The agentic AI platform Ema introduced the concept of agentic business transformation (ABT) last year, in partnership with HFS Research, to fill a gap in the existing vocabulary surrounding AI agents. According to Surojit Chatterjee, CEO and founder of Ema, ABT represents a fundamental shift from previous digital transformations by integrating AI agents at the core of organizations.

Prasun Shah emphasizes the importance of redesigning the entire organization, including the operational model, workflows, decision-making rights, and performance management systems, for AI agents to become true value creators.

According to Ema, ABT encompasses three fundamental pillars: the organization's technology stack, its workforce, and the metrics used to evaluate success. These elements are crucial to ensure that AI agents are not merely one-off tools but active participants in value creation.

AI Agents as the Connective Tissue

The first pillar of ABT is the technology stack. Chatterjee explains that current infrastructures, designed for human-centered processes, must be rethought to integrate AI agents operating at machine speed. These agents should not be viewed as just an additional layer but as connective tissue that coordinates complex tasks and interprets data from multiple applications.

To succeed in this integration, companies must adapt their technology stack to enable AI agents to make better-quality decisions by simultaneously accessing multiple datasets and applications. This would reduce the time to bring new business requirements into production from several months to just a few days.

Leaders must therefore adjust their technology stack to elicit better-quality decisions from AI agents, prioritizing simultaneous access to multiple datasets and applications to develop tacit knowledge. "Organizations that make this architectural change become truly more adaptive," says Chatterjee. "When a new business requirement emerges, you don't wait six months for a software vendor to build a feature. You configure an AI employee using natural language and connect it to the systems it needs. The time from business workflow to production drops from several months to a few days."

The Workforce, Redesigned

With the expansion of AI agents, companies need to rethink the dynamics of their workforce. Traditional hierarchical structures, inherited from the industrial era, must evolve to incorporate hybrid teams composed of AI agents and human employees. Managers will need to handle new responsibilities, particularly concerning trust, explainability, and psychological safety.

In a workforce that blends AI agents and human employees, managers will be freed from many execution tasks but will need to take on new responsibilities related to managing hybrid teams. Managers "will need to be able to address issues around trust, explainability, psychological safety, and even status dynamics" to navigate the new tensions that may arise in a hybrid workforce, explains Shah.

The impact of agentic AI on existing workforce structures goes far beyond the management layer. Organizations will need to act quickly to modify recruitment, retention, and compensation to adapt to these changes.

From Production to Outcomes

Traditional success metrics, focused on production, are becoming obsolete with the integration of AI agents. Chatterjee emphasizes that companies need to develop new metrics centered on outcomes rather than individual deliverables to assess the real impact of AI agents.

A concrete example is that of an Ema client who tripled their return on investment by revising their metrics, focusing on outcomes such as the percentage of contracts reviewed without human intervention.

To correct this, companies must develop a new set of metrics that focus on outcomes rather than production. This means metrics on broader benefits or changes achieved, rather than on individual deliverables.

For instance, when one of Ema's large clients revised their own metrics, shifting from tool metrics like cost per request and AI accuracy to outcomes like the percentage of contracts reviewed without human escalation, the measured return on investment from agentic AI tripled within two quarters. These changes meant that "this client stopped building one-off solutions in high-volume, low-complexity workflows and began deploying AI employees where the value of the outcome was highest," says Chatterjee.

Integrating new metrics may also require a complete reconfiguration of reward and talent management processes, as well as accountability and ownership within organizations, Shah notes. In human-AI teams, for example, while ethical and fiduciary responsibilities likely remain with human employees, operational accountability will become much more diffuse to reflect the systemic role of AI agents.

This shift will raise new questions that leadership teams will need to examine, Shah adds. They will need to consider: Who is responsible when an AI employee makes a mistake? What happens when AI and humans disagree? What protections must be put in place to safeguard customers?

Laying the Groundwork for Systemic Change

Systemic change is gradual. These are complex questions that experts continue to explore. But by initiating an internal dialogue on the fundamental pillars of ABT—the workforce, the technology stack, and the metrics by which success can be evaluated—leaders can lay the groundwork for an organization better prepared to embrace AI agents at a systemic level and begin to bridge the gap between their ambition and execution.