My Experiences With Data: Personal Topics Of Interest

I am a recent university graduate currently working as a data analyst. Personally, what I find is most important to me is growth. If I am in a position where I am not growing as a person, both professionally and privately then I struggle. Whilst this does not mean every task I undertake must be a new task that will add a new skill or technical competence, it does play a big role in my drive and motivation. If I find that I am “stuck in a rut” or have no perceivable room for growth in the horizon, it will be the prime motivation for me to seek out and change the circumstances that I am in.

Data is perhaps my most significant field of interest. I dedicate a large portion of both my free and professional time to data and its applications. This can be from pure data science and analytics to modern ethical data concerns. Consequently, I find most data tasks invigorating and insightful. I often find very mundane and boring tasks such as data labelling and tagging, (which many of my colleagues and peers consider repetitive)  absorbing and engaging.

An exceptional passion of mine is being able to find new and unique data that relates to the problem at hand. During my time at Explorium this was my specific task to enrich clients data. My role was to be able to find additional data to further leverage and improve upon the predictive model the data science team had built for the client. Say for example it was a fraud model to detect a fraudulent transaction and the client would have a given set of fields related to that specific transaction, then being able to provide new and previously unseen data would drastically improve the performance of the model.

Another considerable area of interest is database design and implementation. Too often is a problem associated with a corrupted or poorly designed database. Back in early 2016 when I started at I Ellinson & Co, the company was currently still using paper ledgers to keep track of all its properties. If a tenant wanted to know how much rent they were in arrears, it would require bringing out a number of large paper ledgers spanning many years and manually calculating the rent owed. A clear need for a database was evident for the huge portfolio of over 500 properties (many with multiple tenants). Over the following two years at I Ellinson I successfully digitised each tenant and property. 

This process was not without teething difficulties. Perhaps one of the biggest challenges was deciding one when to complete the switch. This process turned out to be far more difficult than simply picking a date. Say for example some works required to be carried out only where invoiced after the date which was chosen for digitisation, then this property which the works were carried out on is now showing a negative balance, as no income (from rent) has been reported. Subsequently, this task required a full time database manager to carefully integrate the portfolio and ensure it all nicely reconciles on the bank balance. 

Lastly, being able to translate data into easy to understand actionable results is another area of interest. This is a task that many people struggle with in terms of bringing down high level and abstract problems and applying them to regular business solutions. During the time at I Know First this was a very large part of my role. Furthermore, after speaking to many small business owners and explaining to them the issues and problems that I deal with on a day to day basis as a data analyst, they very quickly realise that these are the very same problems that they themselves deal with in their business. However, it wasn’t until the problems and solutions that I deal with as a data analyst had been distilled down and carefully explained in layman terms that enabled them to make this connection.

Finally, in terms of future growth and career direction I am looking to continue solving the same problems but ultimately, be able to find small but meaningful improvements to existing solutions. Personally, I have no grand goal of building a completely new algorithm from the ground up that is able to resolve some big problem. I believe in the benefit of small improvements to current existing methods are perhaps more impactful and past experience has taught me that it will most certainly be more readily adopted into existing frameworks and pipelines.