Ra'ad Costello-Mahmoud

Kromek

In my roles at Kromek, I have developed data pipelines and neural networks to tackle challenges ranging from localising explosive threats in aviation security to identifying novel pathogens from metagenomic air samples. As a now-Senior ML Engineer, I have owned the entire end-to-end lifecycle on several projects from data gathering and curation, architecture design and implementation, and deployment and maintenance. Outside of these projects, I have also made it my goal to gently cultivate an understanding of deep learning concepts and capabilities in the wider company, motivating the hiring of additional data scientists and machine learning engineers to our growing team.

Nissan

During my PhD, I interned as a Junior ML Engineer at Nissan, at their UK Sunderland site. In that role, I developed novel computer vision algorithms for automotive bodywork defect detection. The procedures I developed now form integral steps in defect detection systems used by the company worldwide, to ensure optimal manufacturing quality.

JAXA, Sagamihara

During the first year of my PhD, I had the opportunity to travel to and work at the Japan Aerospace Exploration Agency site in Sagamihara, Tokyo, for 3 months. In this time, I built friendships and connections with local researchers which survive to this day. In this invigorating research environment, I also made the first steps toward the ground-breaking unified spectral-timing model which would form the basis of my thesis 2 years later (see here).

The CEA at Durham

From 2016-2020, I undertook my PhD on the computational modelling of the physical accretion processes around black holes under the supervision of Professors Chris Done and Tim Roberts, at the Centre for Extragalactic Astronomy in Durham. This 3-year project consisted of the development of novel computational models for the joint energetic and timing properties of black holes. Under the hood, these models incorporate Monte Carlo methods, fundamental physical principles, and Fourier decomposition, as well as complex analysis and Bayesian techniques. During the course of this doctorate, I produced four first-author published papers in the Monthly Notices of the Royal Astronomical Society (see here), delivered talks at international conferences in Madrid (Spain), Sigtuna (Sweden) and Guilin (China), and refereed several papers for MNRAS.

Ra'ad Costello-Mahmoud