O’NEILL
In some organisations there seems to be as many architect roles as there are architects (enterprise architect, solution architect, information architect, application architect, data architect, and so on). Len Feskins from HP calls this the “What I do” architect syndrome.
Appreciating this, what do you see as the unique role of the enterprise architect in an organisation?
MCCUSKER
The EA role recognises that you need an architect who is skilled in IT and finance as well as capable of providing a solid contribution to corporate strategy, especially in the areas of business strategy/IT alignment and technology roadmaps.
It’s still true that IT needs to provide useful outcomes and services to the business but for many organisations the difference between the business and the business’ IT systems are almost indistinguishable, especially in broking financial services, and also logistics, distribution and manufacturing.
Just recently we had a complex design where we required 18 specialist architects as complex systems require a large number of specialist skills. Overall, specialisation is following the global pattern adopted in a range of other disciplines such as science, medicine, law and construction and by and large (assuming you can resolve the disputes) you end up with a better design and a smoother implementation.
MCCUSKER
How can enterprise architects assist academics and still be able to manage to pay off the mortgage?
O’NEILL
I think the simple answer is that EAs will have to stay in their high-paying industry positions and moonlight to universities. There are various industry-academia moonlighting opportunities such as: industry advisory network’s; research collaborations and projects; guest lecturing; research chairs; and visiting or adjunct professorships. All these are pretty much unfunded positions (and indeed will often require the industry partner to contribute cash and/or in-kind) but they will allow input to a university’s educational and/or research programs.
O’NEILL
There are well-established techniques in system engineering (for example structured analysis and design, architecture-based engineering) that have been practiced in the defence and telecommunications sectors for years, so why is it that these haven’t gained traction in the broader business community?
MCCUSKER
I’m a really big fan of system engineering, however, as a discipline it requires a structured problem in order for the discipline to start to add significant value. Many complex issues in an organisation are just not “nicely” structured. Even defining the problem(s) can take many months before a set of solutions is proposed.
If at all possible, I try to encourage my architects to clearly separate the set of problems from the set of possible solutions. This is so that they don’t inadvertently define the problem in terms of the proposed solution as opposed to the actual problem itself. (Luckily I’ve never done that myself . . .)
Often problems in telecommunication and defence can lead themselves to a structural analysis, given the type of problems that they are dealing with. For issues such as “why is system Y not being used” requires a set of tools outside of structured analysis, until there is an opportunity to use this tool-set.
Write to the Editor at Technology and Business
* All fields are mandatory.