Enterprise Architecture (EA) has long been about the big picture—vision, strategy, and ensuring the technology landscape doesn’t collapse under the weight of its own complexity. Agile, on the other hand, is all about delivering value fast, embracing change, and avoiding big upfront designs.

When these two worlds meet… sparks fly.

Agile vs EA — The Friendly (Mostly) Tension

From Agile’s perspective:

  • “Why are we waiting for an architecture review when customers want this feature yesterday?”
  • “We value working software over comprehensive documentation… remember?”

From EA’s perspective:

  • “If we don’t think ahead, you’ll ship three incompatible solutions that all do the same thing—badly.”
  • “We value sustainable systems over short-term wins… remember?”

Neither side is wrong. The tension exists because both are optimising for different time horizons. Agile looks at now; EA looks at years from now. And in a complex organisation, both timelines matter.

The AI Twist

Enter AI, and the game changes:

  • Agile teams can prototype AI features fast—but need guidance on how to integrate them responsibly.
  • EA can map AI capabilities to enterprise strategy—but needs to move at AI’s breakneck speed.
  • Data governance, model lifecycles, and compliance frameworks become must-haves, not nice-to-haves.

Suddenly, the EA’s obsession with guardrails starts to look… less like bureaucracy and more like survival planning.

Unresolved Questions

  • Can Agile teams stay truly autonomous if AI ethics and governance require central oversight?
  • Can EA adapt its processes to keep up with AI’s rapid evolution without sacrificing rigor?
  • Will “just enough architecture” work when the AI components themselves are changing weekly?

The answers aren’t fully clear—yet.

My Completely Biased Conclusion

If AI continues accelerating the pace of change, context, coherence, and guardrails will become even more important.
And who are the keepers of that? Enterprise Architects.

So perhaps in the not-so-distant future, delivery teams will be saying:

“We value comprehensive architecture over working software—at least until the architecture is perfect.”

(Note: This conclusion may be slightly biased. Written by someone who is, in fact, an Enterprise Architect.)

In the age of Agile and AI, EA isn’t dead—it’s evolving. And if you ask an EA, it’s evolving to become the only role that truly matters. 😉