Agent-based models (ABMs) are the next generation of analytical techniques to explore the behaviour of complex social, physical and biological systems.
However, ABMs are computationally intensive and developing them is a highly iterative process. Furthermore, they require extensive calibration and validation in order to create meaningful and useful models.
This has hindered their usage amongst analysts and adoption by decision makers.
Public cloud services offer the raw compute resources on a pay-as-you-go basis to address these challenges. But how do modellers harness these resources without having to develop and manage cloud infrastructure?
Start and stop compute resources on-demand.
Scale computer resources vertically and horizontally.
Managed underlying cloud infrastructure.
Configure and manage compute environments.
Easily install and upgrade dependencies.
Reuse and share environments.
Run large-scale ABM workflows using a distributed runtime execution engine.
Define workflows using a lightweight SDK.
Agnostic to modelling framework.