With a total population of more than 884,000, Cox’s Bazar settlement density is similar to New York City, therefore COVID-19 spread is a greater risk. Contrary to many SIR (severity-infected-recovery) models, this project – which is based on an agent-based model originally deployed in the UK – uses available georeferenced data to construct a “digital twin” of Kutupalong settlement. The project, led by UN Global Pulse and with support of the Innovation Service, consists of building a digital twin that includes information on geo-locations of key population movement mixing points such as: households, food distribution centers, water pumps/handwashing stations and schools, and the demographic data of over 700 000 inhabitants. The simulation probabilistically determines what agents (refugees) do during the day. Therefore, disease spread and symptom progression are based on demographic characteristics and the inclusion of comorbidities when calculating the likelihood of severe disease progression.