The RUM Methodology

This document aims to provide to a non-expert engineer, somewhat familiar with the work done regarding Trustworthiness in the confiance.ai program, a synthetic and unified account on what has been accomplished regarding Robustness (R), Uncertainty Quantification (UQ) and Monitoring (M) tools in the EC3 and EC4 teams by the end of the program. Additionally, it aims at providing perspectives for developments in programs that will be sequels of confiance.ai such as CSI and IAG Core.