Short latency APIs and data access notarization for AI continuous training and reinforcement

Overview & Synopsis

This research article aims to propose a new method of API (Application Programming Interface) loop latency notarization to prove that APIs loops used to collect data for AI (Artificial Intelligence) continuous training of an application are responding before a critical specified delay. In addition, the data freshness of data used to train or feed the AI models is also notarized to make sure that the IA model is using compatible data with enough freshness. Notarized API loop latency or data freshness can be stored in blockchain to allow to prove the applicability of the models to the industrial use cases like robotics, control command, digital twins etc.for providing the right environment of researching and thinking on such topics.