This love for both learning about the way the Universe works and using the scientific method and data to understand the world around me led me to pursue an early research career in Astrophysics and Cosmology. During my Masters degree and PhD I learnt about the fundamental laws which govern our Universe and contributed my own piece of understanding to the topic of Large Scale Structure in Cosmology. This work revolved around analysing the large scale distribution of millions of galaxies from international sky-surveys to provide further understanding about processes which occurred in the early Universe and how we can leverage these to understand better the cosmological model (understand components like Dark Energy?). This work involved international collaboration, a key underpinning of the application of the scientific method and the ability to articulate complex research to a variety of audiences. People may envision that astrophysics and cosmology involves much of the time looking through large telescopes and although this is sometimes true, the era of precision Cosmology has occurred because of the advancement in technology meaning that the quantity of data to process requires many of the skillsets sought after in the Data Science community (data languages, high-performance computing, advanced statistics).

 

During my PhD I found that although my dream of contributing to scientific understanding in the field of Cosmology had been realised, there was a desire to apply my knowledge to a domain which feels more tangible. This led to my decision to pursue a career in the discipline of Data Science after my PhD. With a group of fellow PhD researchers who had similar ambitions we set about developing the domain knowledge which you are not readily exposed to which in my case was aspects of Machine Learning and deep learning - although during my PhD I saw others working with ML in different parts of Cosmology (Photometric redshifts, GANs for galaxy mock catalogue simulations and more). 

​

My journey to be a Data Scientist culminated in undertaking a placement at BMT Smart who provide a ship and fleet monitoring service to clients in the Commercial Shipping Industry. Within this industry (which has a global CO2 emissions similar to countries like Germany) regulation on fuel emissions is a hot topic. Large commercial vessels during a voyage and lifetime can have many different things which can effect the fuel consumption - weather conditions, hull fouling, onboard operations etc. I worked for 4 months on a project to implement a predictive model for fuel consumption of a ship given these factors to allow for forecasting tools when combined with weather and on-shore engine instructions. The conclusion to the project was ensemble models of random forest regression and fully-connected neural network architectures offered an improvement in predictive power against standard techniques (Polynomial regression) and allowed for easy incorporation on non-linear relationships between features.

​

Since finishing my PhD I have been working in the role of Data Scientist at an EdTech company FutureLearn which is Europe's largest company within this domain and has provided access to higher education courses to over 10 million learners globally. Since starting at FutureLearn I have worked on a variety of interesting applications of data analytics and science to provide data-driven decisions within the company.