The Three Phases of Evolution Towards Agentic Organizations

Organizations are generally structured around specific specializations and a hierarchy on top. This specialization might be on functions, products, geography, or a combination of these elements in a matrix structure. A hierarchy is then established to streamline decision-making processes necessary for resource allocation and strategic direction. Processes operate in top-down, bottom-up, and cross-departmental manners, creating cohesion among various departmental silos.

In digitally mature organizations, each step in the process generates data—either structured or unstructured. Over the past few decades, organizations have endeavored to capture this data and build analytics that facilitate intelligent decision-making. From a systems perspective, this encompasses transactional systems (such as ERPs and CRMs), data stored in Data Warehouses (whether cloud-based or on-premises), and dynamic reports presented as KPIs and Dashboards. The traditional flow of this process is:

Human Input – Transaction – Digital Output – Human Reasoning – Human Decision.

*With the advent of Machine Learning predictions, the final stage can also be supported by AI.

With Agentic AI, the flow will shift to:

Agentic Input – Transaction – Digital Output – Agentic Reasoning – Human or Agent Decision.

Humans will engage in Management by Exception based on the recommendations provided by the Agent responsible.

Organizations embarking in the Intelligent Process Automation journey will go through various organizational stages- evolutionary approach.

In Phase I, we anticipate that organizational structures will remain unchanged, but they will incorporate Agents into their assets. Agents are likely to emerge in digitally mature organizations—those where all major processes operate on enterprise software with a significant degree of system integration. These organizations possess a strong digital culture, robust IT department, and a willingness to change driven by competitive pressures or other factors. They will view Agents as integral to their Intelligent Process Automation (IPA) initiatives with strong Human in the Loop (HITL) engagement. Initially, Agents will be developed for the most straightforward use cases within Finance, HR, Marketing, Customer Care, and IT departments. Each department will have specialized Agents performing tasks as part of their processes. During this phase, departmental silos will persist, connected only through software applications such as Enterprise Resource Planning (ERP) and IT Service Management (ITSM) systems and non-Agentic technologies (APIs, RPAs etc.).

In Phase II, as technology advances, we will deploy Agents to assist users across various functions. For instance, there may be one Agent responsible for handling all employee queries ranging from HR to IT that will  coordinate several specialized Agents. During Phase II, we will implement cross-departmental orchestration of Agents. Processes will become connected through Orchestration Agents that will predict and plan the actions of the specialized Agents. This phase will see the breakdown of organizational silos and the creation of a process intelligence foundation. Humans and Agents will collaborate within the same processes, with humans serving as the escalation point for Agents or when Human to Human  interaction is necessary. In this phase, critical questions will arise regarding the ownership of the orchestration. For example, who will oversee and manage conflict resolution within the agentic process? Is it an Operations Officer, a Business Improvement Officer, the leading Department Head?

We anticipate that in Phase III, organizations will develop a new type of structure centered around teams composed of both Humans and Agents. With the advanced level of automation, employees will primarily focus on strategic planning, project management, creative tasks, competitive analysis, and other initiatives that require significant interaction, planning and creativity .

In Phase III organizations will run on hybrid intelligence (AI and Human) in the same way as now they are running in hybrid environments (physical and digital). Organizations will held hybrid meetings where employees will be accompanied by their specialist Agents, who will follow up on actions and tasks. Human input will be provided to Agents for planning, executing tasks, and coordinating activities among humans until the next meeting. This environment of agentic and human interactions could potentially lead to a scenario where humans may never or rarely interact directly with the “old” software applications but only via Agents.

Organizational Evolution towards Agentic Organizations

Organizations will compete based on the efficiency of their hybrid environments, how effective are their specialized Agents, and how effectively Agents support their employees. Their focus will be on how employees engage in meaningful human interactions that add significant value to the organization. Finally, new organizational cultures will emerge that will be reflected in the way that Agents will interact with humans. Organizations will create their own “Agentic language tone” and pass their own “value systems” through A2H (Agent to Human) interactions. Eventually Phase III organizations will use Agents as their change enablers…

But that doesn’t mean that Agentic AI will automate all processes. Similar to how craftsmanship labor and small specialized furniture producers still exist, companies in the future will automate parts of their processes while retaining Highly Emotional, Complex, and Critical tasks for humans. Future competition will hinge on the effectiveness of Intelligent Process Automations (IPA) and the strategic utilization of human intelligence to create a competitive advantage.

USE CASE

A field engineer is attempting to repair a machine but lacks essential technical information. He contacts the Intelligent Library Agent for the necessary details. Unfortunately, the organization’s library does not possess the required information, prompting the Agent to reach out to the suppliers’ Intelligent Library Agent for assistance. The supplier’s Agent verifies the customer's enrollment in the service and promptly provides the information to the Agent. This scenario can be extended to trusted Agents throughout supply chains with result Agentic Supply Chains.