A company’s executives should not need an EA to walk them through technology. What the board room needs is a recommendation.
Do this, for this reason, and here is the price of getting it wrong.
EA Knowing products is not a recommendation. But, EA Judgment can, and the C-Room has to begin asking, before any design work begins. (see Vocabulary for the terms used).
Starting with the outcomes
The EA’s first advice to the C-room, should not be: “how do we build it?” Instead: “what is the company trying to achieve, and is this the right thing to build?”. Further to that EA should ask the board:
- Name the business outcome wanted.
- Describe where the company stands today and where it must be.
- Find what blocks the way.
- Plan across technology, process, people, data, security and operations to remove the block.
The AI Context: “Let’s use AI” is not an outcome. An EA starts from the business result wanted, and only then asks whether AI is the way to get it.
Weighing the options
Only after that does the central work begin. The EA lays out the candidate paths, estimates what each costs, lists its risks and what it gives up, and follows each choice through to its consequences.
Then the EA should commits to one path and explain why.
The AI Context: the options include “do not use AI here” and “use something cheaper than AI”. Each AI option carries its own costs (compute, data preparation, people) and risks (wrong answers, data leaks), and they must be weighed like any other.
What the EA recommendation rests on
It does not rest on novelty, on how interesting a platform is, or on how many product features the EA has memorised.
It rests on the business, its constraints, the risks and the alternatives, weighed together. The result is an answer that can be defended in plain words to people who are not technical.
The AI Context: AI tools change every few weeks and every vendor claims to be the newest. A recommendation based on novelty ages as fast as the tool. One based on the business and its constraints does not.
True but. EA knowledge serves the judgment
EA technology knowledge is necessary, but it is an input. The decision is the product, and making it well is what separates an EA from a well-read technician.
This is what executives are actually asking for: complexity reduced to a direction they can act on with confidence.
The AI Context: AI can now supply much of the technology knowledge on request. What it cannot supply is accountability for the decision. That stays with the EA, which makes judgment the part of the job that remains.
Vocabulary
- EA (Enterprise Architect): the person who decides how a company’s technology, processes and people fit together to serve its goals.
- Executives: the people accountable for the company’s direction and money.
- Trade-off: what is given up by choosing one option over another.