Part III The faster you learn is money

DEVOPS an emergent Intelligence Business Architecture

  • Arthur de Snaijer
  • 17 november 2016

An innovation awareness story on who is learning from who, you from the computer or the computer from you. This is the third and final blog in a series of three. Click here to read the first blog, click here to read the second.

Architecting like Starlings

While software has architecture, the same applies with business. But are enterprises that are moving towards economic hubs becoming more sustainable to change? In a new circular network of interchanging business relations each party is growing and is also trying to be resilience or Anti-Fragile. How does an enterprise nervous system react like sparlings in a high competitive environment? When you compare yourself with a staling you’re personal flight is aware of a better business model that strikes-back. Due to this you change flight direction or disappear. Not changing is a sparling that hits the wall. Wanting to change you need to react faster, so you are able to change your business model while observing the market environment? How to do this. You could DevOps an intelligent nervous AI/Human business core learning model. Or from a design-thinking perspective; integrate API-data driven events and continuously evolve you analytics learning system that API’s with the world?  

To give you an idea of what I mean I tried out a tree experiment and share my learnings in this blog.


My data driven tree experimentation

I started with a simple tree experiment. In my small office-cubicle a tree plant needs to watered each time. Having other priorities, I really disliked fetching water. This is because the water tap is on the other side of the building.  After fetching water two time time’s I realized that some automation is feasible. I life hacked a construction (prototype) to provide my plant with the water & nutrition’s it needs. Having automated the process I received IoT sensor data about drinking habits. I used this feedback to recognize if the plant was growing-up healthy? This technique can also be read as feedback to use do determine business process maturity. All sensor micro data is recorded into a big data environment. Having plant learning data, I wanted to know which place in the room my plant would be happy. Sunny or not spot, some shade or a lot.  To answer this question I needed to science and crunch some numbers. How to do this? 


Not muscle memory, but a fast learning business case

Unfortunate not knowing all the skills to data driven science the heck out-of-plants, I asked a colleague analytics professional on approaching this problem. She unfortunate due to high–demand was busy with other clients. Not having anybody to explain me how, what are my learning options for this experiment?. Could I ask the market (buy-a-pro), or is learning myself better? A small time investment. I started browsing in-house @Ordina Academy, the Enterprise knowledge center (wiki) and the Internet, on devops, ai, data, analytics training. Here is what got my attention:

Not having all the time in the world I needed a self learning nano master. From the things that got my attention I selected just-in-time/minimum valuable knowledge for this new skill experiment. Picking knowledge cherries. This is like a sparling eating flies during flight. Ordina is sharing intelligence on the enterprise wiki, so I can learn and move our enterprise thinking faster forward.  Pretty good that my colleagues are pushing experimentation potential in the right direction. Knowledgeable Professors, also do this in a different marketplace or not?

Invest in high performance team learning

When in a living company a person is learning from the computer and a computer is learning from you. You can enhance this towards a team level, the high performance team (HPT). Above the left video is an example in how interoperability between humans and machines can grow. The right is what to expect when business will drive on AI data learning models, and what a service or product HPT can expect to come.


Returning to my blog title an emergent intelligent architecture. Accelerating learning is a business design  in which a people way of thinking use data & learning parallelism to scale.  Business services and products benefit from recognizing personal, team, enterprise and chain value learning. Using my tree nutrition model, I can now scale this learning model data using a Graphcore. When a plant anomaly occurs – e.g. the nice weather changes rapidly – my AI-nervous engine observes and acts by enhancing the office artificial lighting.  Always in a sunny spot.  Graphcore’s key differentiator is that its technology will offer not only data parallelism but also instruction-level parallelism. Transforming this in business language. Economically think on what happens when “learning people” on a HPT scale faster than business competitors. That’s = Money.