6 years, 10 months ago

Algorithmic Enterprise Architecture

Can some or all ‘design’ decisions we make as Enterprise Architects be automated via collection of datapoints, algorithms and assessment against patterns of architecture?

I read an article the other day in Wired magazine about jobs being consumed by algorithms e.g. Taxi drivers and google self driving cars. It made me think about what I do as an Enterprise Architect and what elements of an ea role could be subsumed by an algorithmic approach. So I thought i’d write a brief post to nail down a few initial thoughts.

Some assumptions:

1) When we engage with stakeholders to understand the baseline architecture we are effectively capturing data points and their relationship to one another.

2) When we map different layers of an architecture we are merely segmenting some datapoints.

3) When we map views of a target architecture and the transitions to it we are taking our understanding of existing datapoints and mashing them against new ones around vision, principle, drivers, goals objectives, constraints

4) Whether they know it or not, all organisations are a collection of (known or yet to be discovered) patterns or anti-patterns. 

Assuming the previous statements are true (big assumption but lets pretend for a minute). What if there was a tool that consumed datapoints of an Enterprise Architecture and datapoints of the vector of the future state and in so doing created simulations and recommendations on the optimal target architectures for the organisation?