This post is written as both a message of hope that there is no such thing as a single career path and as a (self) reminder not dwell too long on past career decisions. Something I am often prone to do.
How and why did I decide to become a Data Scientist?
If I was in a dark mood and being extremely flippant, I would probably tell you it was because I failed at a long list of other careers before the opportunity arose to go back to school to study Data Science.
ArchitectStructural EngineerRadio ProducerStatistical CompilerDeveloperDigital Services ManagerDigital DirectorChief Technical Officer- Data Scientist
However the reality is that almost all of the steps / turns in my career path have helped move me in the direction of becoming a Data Scientist, even if I may not necessarily have been completely aware of it at the time or made the next step with that destination in mind.
School, the journey begins
I vividly remember the intense pressure about choosing a career. I felt as if I only had one chance to choose and, whatever choice I made, would have to last me the rest of my working life.
I’m not entirely sure why I felt such pressure, if only I had drawn on my father’s experience, I would have come to realise that his dreams of being a pilot in the Royal Air Force had been quashed after his eardrums burst on a training flight and he subsequently had to reassess, firstly choosing to become an insurance actuary but later switching to becoming an accountant. He would always tell me “whether business is up or down, you’ll always need an accountant”.
Whilst there was certainly pressure, it didn’t stop me making a decision as, from around the age of 11 I was obsessed about becoming an Architect.
Far from believing that I was destined to become the next Frank Lloyd Wright or Le Corbusier, I was driven by the desire to design everyday, residential dwellings or redesign existing dwellings so that they made better use of the living space.
It was at school that I took my first step, albeit unknowingly and somewhat forcibly, towards a career in Data Science (instead of Architecture). I was unable to mix my choice of science subjects with art at A-Level as the timetable “wouldn’t allow it”. Unable to explore my creative side, I doubled down on the science, choosing to study Mathematics, Further Mathematics, Physics and Chemistry instead.
Data Science Lesson: Studying Pure Mathematics was where I first got an understanding of statistical methods.
Data Science Skills: Statistics
University, and “giving up on my dream”
I progressed to University having got my A-Levels. Admittedly, I did not do well enough to get my first choice of University but a well known architectural school nonetheless, The Bartlett School of Architecture at University College London.
It was a disaster.
I struggled from the first day in part because I couldn’t match the level of creativity required (I should’ve done that Art A-Level after all), but also because I wasn’t used to the unstructured and self-driven nature of learning at university (something I’m coping much better with on the Flatiron School Data Science course).
I persevered throughout my first year but at my end of year review with my tutor it was suggested that perhaps Architecture wasn’t for me and I should consider changing my course of study…. I was devastated.
I loved living in London and really didn’t want to think about taking a year out to take time to reflect properly and consider all my options. I had also begun to make friends and didn’t want to have to start all over again at a different institution in a different city.
After a couple of meetings with Professors from the relevant faculty, I decided to change course to study Structural Engineering.
It was a hard course, with significantly more time spent each week in lectures compared with the Architectural course which had been dominated by projects. But I was back in my comfort zone studying subjects such as Mathematics, Structural Design and Fluid Mechanics.
Data Science Lesson: Studying Engineering introduced me to the concepts of programming and computational modelling.
Data Science Skills: Unix programming, Statistics
Into the job market, and changing direction again
I graduated in 1996 and at that time there were no jobs to be found in engineering. Out of our entire graduating class, only 2 of us managed to secure jobs within the engineering industry so it seemed that the path of my career would change direction again.
I wasn’t really sure what I wanted to do. I was playing guitar in a band at the time and so considered becoming a full-time musician. I also did a couple of internships with the aim becoming a Radio producer.
I finally got a job as a statistical survey compiler working for the Chartered Institute of Public Finance and Accountancy (CIPFA). You can read a little more about my first “proper” job in another blog post here. It was here that I really started to understand the value of using data as a way of making decisions and affecting positive change.
One of the services that I worked on for CIPFA, provided local authorities in the UK with a range of statistical profiles that compared one local authority with a group of other, “nearest neighbour” authorities. This benchmarking service emphasised the importance of increasing service performance to the general public rather than simply cutting costs.
Data Science Lesson: Not only did I get the opportunity to learn and improve my coding skills, I got a great understanding of how data could be collected, manipulated and examined to improve both financial and non-financial performance.
Data Science Skills: Statistical modelling techniques, MSSQL, ASP, VBA and Excel.
Building my alternative career
After 10 years at CIPFA, and for various different reasons, it was time to move on…
D&AD, is a “non-profit advertising and design association, investing their surpluses into educational programmes inspiring the next generation of creative talent and stimulating the creative industry to work towards a fairer more sustainable future”.
The business model at D&AD is somewhat unique, with the majority of income being made via the D&AD Awards, often described as the Oscars of the design and advertising industry, and which used to come into the business in a very short 2 to 3 week window in February each year.
The trouble was that the business had no clear visibility about how much money was coming in, how this compared to previous years and expectations and the total amount would vary year-on-year.
I developed a forecasting / reporting system that not only tracked, minute-by-minute the income that was being generated as Awards entries were submitted but also compared the performance with previous years and ultimately predicted a target income for upcoming years’ Awards cycle.
Data Science Lesson: Using data to predict and solve business problems.
Data Science Skills: Client communications, statistical modelling techniques, MySQL, Salesforce, Python, Git.
Joining an Agency
After almost 10 years at D&AD, it was time to move on again…
JP74 was an “emerging technology consultancy”, making the most of what already exists, and creating what doesn’t through platform integration and bespoke development.
Having multiple clients was something new for me and probably one of the main motivations for me joining JP74. During my time with the company we worked on a diverse range of projects from an integrated CRM and room allocation system for a company specialising in accommodation for International Students in London to an analytics platform for egg production in the UK.
The companies that we worked with as a consultancy all had one thing in common, there was a data problem at the heart of their business. Understanding that data problem would not just mean the success of the project, but the overall business.
Data Science Lesson: Understanding that there is a data problem at the heart of all businesses.
Data Science Skills: Client communications, data analysis techniques, MySQL, SugarCRM, AWS, Python, Git.
Back to school
So, in truth, the reasons behind my decision to become a data scientist were not because I have some sort of natural gift for statistics or programming or as a result of some moment of clarity where I suddenly realised that data science was my calling.
I took the opportunity to go back to school to study Data Science because it brings together the different aspects of my career so far, statistics, programming and data-driven decision making.
At the time of taking each step, it didn’t seem like I was on a specific path but Data Science represents the next logical step to take in the course of my career.
The final project submission for Module 3 of the Flatiron School Data Science course has focussed on creating a classifier model that predicts the functional status of waterpoints in Tanzania and these are the types of projects that I get excited about. Where the manipulation, understanding and modelling of data can have extremely beneficial effects in the real-world.