Brendan Foody describes how organizations use evaluations to measure model performanceEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria. and decide where post-training may help.
He forecasts substantial automation of digital tasks while identifying long-horizon work, large environments and human interaction as continuing challenges. These timelines are his predictions, not established outcomes.
Foody also discusses his company’s agent workflows and the role human expertise may play in measuring and improving future systems.
Watch on YouTube




