Agentic AI : the next silent transformation of systems and organisations

By Simon Georges Kot, Partner – Eleven Strategy

Agentic AI is attracting growing attention. Often presented as a new form of intelligent automation, or as the natural evolution of RPA, it raises both curiosity and questions. However, beyond experimentation and technical demonstrations, the real challenges lie elsewhere : in the ability of organisations to integrate it into their systems, processes and culture. This represents a profound transformation, impacting IT architecture, governance and the very way of working.

Structuring systems to support a world of agents

Agentic AI is not simply a more advanced chatbot, it is a new architectural paradigm. While traditional systems rely on linear and centralised flows, agents introduce a distributed, interoperable and adaptive logic.

Each agent operates as an autonomous microservice, capable of interacting with others, executing complex tasks and adapting to its environment.

To support this new paradigm, information systems must evolve significantly :

• Designing modular and API driven environments
• Rethinking connectivity, no longer between applications, but between cognitive entities
• Preparing for a future of "MCPisation" of information systems, where agents communicate through standardised protocols

This transformation is subtle, yet fundamental : current infrastructures were not designed to support millions of autonomous agents operating continuously. IT leaders who fail to anticipate this shift risk being constrained by architectures that are too rigid to evolve.

‍

Structuring the transformation : the CIO as the future "HR of agents"

As organisations deploy these agents within their processes, the role of the IT function is undergoing a profound shift. In the future, it will no longer manage only technical environments, but actual populations of agents : who does what, with which permissions, responsibilities and under what level of supervision ?

In other words, the IT function is gradually becoming the HR function for agents. This implies rethinking :

• Governance : who orchestrates agents and their interactions ?
• Security : how to isolate, monitor and control autonomous entities ?
• Capability modularisation : how to build reusable building blocks, reasoning, planning, execution, interaction, at scale across the organisation ?

Some organisations are already choosing to develop their own orchestration layers, in order to ensure sovereignty and limit dependency on large language models. This represents a strategic positioning, comparable to what ERP systems or cloud computing once were.

‍

Transforming culture : towards a "no code native" organisation

The agentic revolution goes far beyond the scope of IT. It signals a profound cultural transformation, one in which every employee becomes an active contributor to automation.

Thanks to no code and low code tools, business teams will be able to design their own agents to simplify daily operations : marketing teams will create campaign agents, HR teams reporting agents, and finance teams control agents.

However, unlocking this potential requires preparation : democratising access to tools, supporting upskilling and fostering a culture of controlled experimentation and responsible supervision.

Organisations that succeed in this transition will be those capable of combining architectural rigour, technological modularity and trust in local initiative.

From automation to transformation

Agentic AI is not just about doing things faster. It represents a true paradigm shift : systems that can decide and act, where previous applications were limited to execution.

Data and IT leaders now have a unique opportunity to redefine their technological and organisational foundations, paving the way for a new era of digital autonomy, where humans and agents coexist within an intelligent, scalable and sovereign ecosystem.

Why Eleven ?

At Eleven, we have been supporting large organisations for more than ten years in these technological and organisational transformations. We have witnessed firsthand how they have had to reinvent their cloud architectures, open up their data and integrate generative AI.

Agentic AI is part of this same trajectory of disruption. Our conviction, built through experience, is clear : succeeding in this transformation requires more than experimenting with agents. It requires the ability to structure, govern and embed them within the organisation, so they become true levers of performance and sovereignty.

‍