Contextual AI adds explainability to different stages of machine learning pipelines – data, training, and inference – thereby addressing the trust gap between such ML systems and their users. It does not refer to a specific algorithm or ML method — instead, it takes a human-centric view and approach to AI.