10107849GraphChat: A deductive explainable knowledge graph management system automating complex reasoning with a conversational expert knowledge ingestion and interrogation engineClosedCollaborative R&DInnovate UKOne of the current big challenges in AI lies in balancing scale, speed, explainability and accuracy. Current inductive, or probabilistic methods like neural networks and deep learning offer speed and accuracy in areas such as image recognition and natural language processing but lack full explainability (i.e. the reasoning behind the output) and may be prone to bias and hallucinations. This is problematic in high-risk sectors like medicine and finance, where explainability is key. Prometheux provides a deductive reasoning system utilising knowledge graphs without requiring complex implementation work. This system integrates various data sources for comprehensive data processing and reasoning. This project develops a conversational interface that allows experts to interact with and build knowledge graphs without having to learn a specific programming language. This streamlines processes in complex fields like pharma, law and healthcare.