Usman Islam

My project seeks to build better structured representations of tasks in decision making problems by leveraging compositionality. My research interests are in improving the generality and explainability of decision making systems, particularly in (multi-agent) reinforcement learning.

I chose the STAI CDT as its trust/explainability goals are in line with my ideas and I wanted to benefit from the collaborative cohort-based environment.

Masters Qualification: Integrated masters (MSci) in Mathematics

Work Experience:

  • Research Engineer at Huawei Noah’s Ark (2021-2022)
  • Software Development Engineer at Amazon (2017-2021)