What we do

Our analysts - Alan Jackson

What does your job involve?

My role covers a wide range of tasks from carrying out analysis for clients to in-depth numerical analysis and data mining.

Can you explain how data mining techniques are helping Sumerian clients?

Data mining enables us to take massive volumes of data from our clients, which can involve one hundred million or more individual records. The data often comes to us in different formats and types, termed as "dirty data", which means it’s difficult to make sense out of it until it’s been cleaned. We clean the data first, and with data mining techniques apply advanced statistical algorithms. This allows us to discover significant patterns that can be then followed up by more in-depth manual methods of analysis. It also means we can build a strong foundation of trusted data that our clients can reuse again and again, enabling them to get the answers they need for many different initiatives.

Can you tell us about some of the analysis you have been involved with recently ?

Not long ago, I worked on a piece of analysis for a large investment bank who was looking to relocate and reorganise its datacentres as part of its overall cost reduction strategy. As well as trimming costs, a key consideration was ensuring that the performance of its critical trading applications would not be put at risk. Because of this, the IT team needed to provide assurances to the business that end-user performance and links to external trading systems were not going to be impacted after the move.

How did you approach the analysis?

Firstly, we captured packet traces from network probes which were deployed at various points within the existing datacentre. We then modelled and analysed the data to build a baseline picture of the applications’ current performance levels. From this position, we were then able to apply scenario modelling and understand whether the applications would be impacted by the relocation. The resulting analysis provided the bank with the facts it needed to ensure its anticipated cost savings would not expose the business to any unacceptable risk or impact existing services. You can read more about how IT Analytics helps achieve successful datacentre consolidation in our white paper.

Why do you think that IT Analytics provides such a good approach to decision making?

In a nutshell, because it’s based on hard facts. When making important decisions like the example above, some businesses can fall into a number of traps that can distort the true picture, and so end up making poor decisions that aren’t right for their business. For example, they might only have access to partial data or have people that are too close to the problem – and they just don’t see the wider picture. Or at worst, some people might have a personal agenda to push a certain decision or empire build, and again this isn’t thinking in the business’ best interests. By gaining facts based on real data to make important decisions, as we do in IT Analytics, that kind of old school decision making is cut to pieces.

What’s your background and interests?

Well I’m originally from New Zealand and studied Mathematics at the University of Canterbury in Christchurch, specialising in Numerical Modelling. Previous to Sumerian, I worked on large-scale data warehousing projects for a number of major financial services companies. In terms of interests I’ve always been a computer person, and a studious follower of the "geek culture" whatever that is! Some of my favourite sites to indulge the geek in me are:

• Ars Technica - technology news site that provides a unique combination of technical savvy and wide-ranging interest in the human arts and sciences.
• XKCD - in their own words: "A webcomic of romance, sarcasm, math, and language"
• Engadget - excellent Web blog on the latest gadgets and technology news.

*Please note that Sumerian is not responsible for the content of external sites.