Our co-founders met at University College London during their PhD research. Their research work concerns applying fully automated artificial intelligence solutions to various control, design and optimisation problems in core, metro and data centre communication networks. They are interested in applying these fundamental techniques to the general area of logistics problems in networked systems.
What we do
We aim to build a platform that can be used to solve a variety of logistics problems in networked systems. Found in a range of sectors from data centre management to production line optimisation, many logistics problems in networked systems are NP-Hard combinatorial optimisation (CO) problems. This means that no computationally tractable solution exists, and that high levels of domain expertise are required to hand-build heuristics that find sub-optimal solutions which are not robust under dynamic problem scenarios. For example, the optimal means of distributing a COVID-19 vaccine may change from city to city, and may also change dynamically over time as various groups of the population are vaccinated. Ideally, a new solution would not be required every time such a change is observed. Our research has harnessed recent advances in reinforcement learning to automatically solve these types of problems in data centre networks more flexibly and efficiently than other hand-tuned approaches. Our methods have also been shown to be applicable to a range of other real world CO problems. We intend to create a platform that generalises this technology and apply it to a variety of application areas.