Mind Construct

Mind Construct developed the ASTRID-system, which is a knowledge-graph based system capable of full unsupervised learning of cross-domain knowledge (sparse datasets) without the need for information tagging. The system dynamically maintains its knowledge graph, even based on ad-hoc availability of new information.

They use their in-house developed models and algorithms for semantic analysis of training-data and continuous training of their Common-sense knowledge base, in addition to training projects for our clients.