I am a mathematician with expertise in optimisation and operations research. I am particularly interested in problems that are difficult to solve directly by mixed integer programming (MIP) solver. To this end, I have developed extensive experience in decomposition techniques, such as Dantzig-Wolfe reformulation and Benders' decomposition. It is this expertise that brings me to my current role at Quantagonia, where I apply decomposition techniques to develop hybrid computing algorithms to solve optimisation problems as quadratic unconstrained binary optimisation (QUBO) problems. The aim of my work and research is to combine the power of classical and quantum computers using decomposition techniques to solve challenging optimisation problems.
My academic journey
Prior to my transition to industry, I was on the academic career path. After my PhD in Mathematics, I completed a three year posting at the Zuse Institute Berlin (ZIB). Following that, I had the wonderful opportunity to focus on my own research after being awarded an EPSRC Fellowship in 2017. I first took up this fellowship in the Management Science department of the Lancaster University Management School. Then, in making life decisions, I decided to finish that fellowship at the University of Exeter. This led to me becoming a Senior Lecturer in the Department of Mathematics at the University of Exeter before I departed to join Quantagonia.
A little more details about my research background. As part of my EPSRC fellowship, and most recently, I have been working on general algorithms for decomposition methods. A major achievement of this research was the development a Benders' decomposition solver, using SCIP and GCG. During this time I also developed software for optimising brewery production planning and scheduling. You can see a talk on this here.
A postdoc in Berlin
At ZIB I was part of the SCIP development team. In this role I was engaged in projects that include the development of the internal parallelisation for SCIP, investigating the use of decomposition for the simplex method (implemented in SoPlex) and the development of a general purpose Steiner tree problem solver (SCIP-Jack). This time helped me gain experience in high quality software development that is necessary for developing and maintaining a state-of-the-art MIP/MINLP solver.
PhD working on airline planning
My career in optimisation and operations research kicked off with the awarding of my PhD in mathematics from UNSW in January 2014. My PhD applied the technique of recoverable robustness to airline planning applications. More generally, I formulated and solved integer programming problems that attempted to make the allocation of aircraft (and crew) to flights less susceptible to disruptions.