CV

Education

• Ph.D in Artificial Intelligence, Queen Mary University of London, 2024
  • Thesis title: Multi-task Reinforcement Learning for Adaptive Agents
• BSc in Computer Science, University of Essex, 2018, 1st Class Honours
  • Main modules: Data Structures & Algorithms, Object-oriented programming, Functional Programming, Artificial Intelligence, Computer Vision, Robotics, Game Development, Information Retrieval, Natural Language Processing, Databases, Networking, Computer Architecture, Web development
  • Programming languages: Python, Java, C++, C, Haskell

Work experience

• QinetiQ - Research Assistant; 2023 October - 2024 April:
  • Developed a new variant of the Rolling Horizon Evolutionary Algorithm to control multiple units simultaneously (MURHEA) and implemented a fast simulator of the target environment that was used for decision-making.
• Tabletop R&D - Research Assistant; 2023 March - 2023 June:
  • Implemented a digital twin for a published board game in the Tabletop games framework and ran large-scale testing using various AI methods to get better insight into the game’s balance and strategy space based on various questions that the designer had in mind.
  • The results have been demonstrated at the UK Games Expo 2023 with a report available online
• Microsoft - Research Associate; July 2020 - December 2020:
  • Worked on Project Malmo, a Reinforcement Learning benchmark built on top of the popular video game Minecraft. The framework supports both single and multi-agent RL research with lots of customisable settings. Malmo presents lots of research opportunities, but it has been considered difficult to use and get started with. My job was to lower the entry barrier for new users by improving the Python API and creating examples of how to use the framework.
  • At the end of the project the new framework was presented as a Microsoft Research Webinar and also as a tutorial at the 2021 IEEE Conference on Games
• Defence Science and Technology Laboratory (DSTL) - Research Assistant Programmer; 2019:
  • Implemented a JAVA version of the Pommerman framework with various Statistical Forward Planning agents, including Monte Carlo Tree Search and Rolling Horizon Evolutionary Algorithm. As Pommerman presents various challenges (Partial Observability, Multi-agent, long-term credit assignment) the win/loss ratios were not informative enough to compare the agents, which required further analysis of their behaviours.
  • A paper detailing the analysis of the agents and the framework has been published at AIIDE 19’ and the framework has won the Best Artifact Award.
• Queen Mary University of London (QMUL) - Demonstrator/Teaching Assistant; 2019-2023:
  • Game Artificial Intelligence (Classical Game AI; Statistical Forward Planning, MCTS, Reinforcement Learning)
  • Multi-platform Game Development (Unity)
  • Data Analytics (Python; matplotlib, sklearn, scikit-learn, pandas, numpy)