News & Resources

How can Machine Learning help the Enterprise Architect?

In Software & Cloud Economics

Machine learning can crunch through your corporate data and unearth a few home truths 

It stands to reason that organisations which let data fuel their business decisions are likely to be more successful. So how can machine learning take data and use it to manage and organise enterprise systems? 

Digital transformation has the potential to overhaul the business practices of many organisations, but the key to its success is data and how it is used to effect change. Depending on who you believe, around 90 per cent of all data that has ever existed has been created within the last two years. Yet, only one per cent of that data has been analysed, according to McKinsey.  

This data has the power to help change organisations for the better. Organisations that enable data to help in making business decisions are often more successful as a result. 

Leaving business improvement plans to trial and error and gut feelings often fare worse than those that embrace data.

However, there is a problem. 

Enterprise architects can use some data to create models, but the data within them can and will age, meaning conditions that models tackle will have changed during the planning and implementation phase of a transformation. This means digital transformation threatens to make inflexible architectures and data models obsolete before they have even begun. 

The way to tackle this is through machine learning. It can take the guesswork out of essential business decisions. The insight from machine learning can assist organisations to make better predictions by continuous learning and adaptation for models using real-time data. 

Why use machine learning in the enterprise?

Machine learning is relevant to many industries today and is having a huge impact on enterprise architecture. The technology is being used more and more in enterprises to optimise back-office and consumer facing processes and systems. It will also assist in managing and organisation complex enterprise systems. 

For the enterprise architect, machine learning will be all about creating models that can use algorithms to ingest large quantities of data, something that is beyond manual processes for a human, in order to provide insights and actionable strategies. From data it can learn what is happening within the organisation and see the patterns that maybe invisible to humans. 

While machine learning can be thought of as particularly clever and complex pattern recognition, enterprises can use what is called “deep learning” to solve problems that have been beyond the reach of humans.

Using a framework for machine learning

It is an important caveat to mention here that machine learning cannot be used in every situation. Therefore, it is essential that a framework is in place so that the enterprise architect can provide guidance to the rest of the organisation about the problems needing to be solved. 

Firstly, what are the problems that a business is trying to solve? For enterprises, those that are data intensive are good areas for the application of machine learning. 

Secondly, machine learning is not just about automating manual processes. This means processes that are ripe for machine learning are connected to decision-making have a cognitive aspect, too. 

Lastly, enterprise architects should discover those areas where decisions need to be made in real time and where machine learning can be used to decrease response times. 

Therefore, the operational challenge for enterprise architects when dealing with machine learning is to work out where you can increase the intelligence of a system so that enterprises can make faster decisions to gain a competitive advantage.

So, when embarking on using machine learning within the enterprise, architects need to ensure that the data going into training such systems is of high enough quality in order to obtain the necessary results. Get that right and your investments will pay off quickly. 

Crayon’s Artificial Intelligence (AI) & Machine Learning (ML) Practice, which operates from our European Centre of Excellence, already supports clients around the world across multiple industry verticals. To discuss any requirements you may have, please get in touch.