For many years, enterprise architecture focused on organizing systems.
Where data lives.
Where processes run.
Which platform serves as the system of record.
Which one coordinates workflows across the enterprise.
These questions shaped how architects designed integration patterns and defined boundaries between systems. These are still important today. However, there is a new element in the picture: intelligence.
With the rise of large language models and AI agents, architecture is no longer only about systems, data, and workflows, the new design has to account for the use of intelligence across the enterprise.
Some platforms are starting to embed these capabilities directly within their ecosystems, like ServiceNow, Salesforce, and Workday, Intelligent assistants, AI-driven workflows, and embedded copilots are becoming part of the platforms.
This creates an architectural consideration.
When AI is embedded within a system, it can enhance that platform’s workflows and data processing. But this raises an enterprise question: how does the rest of the ecosystem benefit? The emerging model points toward intelligence services that operate both within platforms and outside them, allowing capabilities to be reused and orchestrated across multiple systems.
The challenge for enterprise architects is finding the right balance. Not only organizing systems and workflows, but deciding where intelligence should live within the architecture.
Jesús Melendez 