SciStack is an AI assisted, transdisciplinary research platform that is advancing the state of the art in applied ontology development, especially for strategic and multi-level C2 system interoperability, and constructing trustworthy AI systems. SciStack applies emerging, and is developing new techniques for automatically generating ontology components, significantly increasing speed and reliability while reducing cost.

SciStack’s foundation is Basic Formal Ontology (BFO), the DoD / IC standard. SciStack has three major components that are supported by a transdisciplinary team and structured process. SciStack’s “Input Sources” supports the ingestion, management and storage of various types of data, both structured and unstructured, as well as reference ontologies and SME’s. The “ontology engine” is the core of KaDSci’s growing contribution to advancing the state of the art in applied ontology development. It is a semi-automated process that process inputs using RAG supported LLM’s to associate inputs with existing ontology terms where possible and creating new ones where necessary. Before finalization the candidate ontology is tested for quality and logical consistency. SciStack has a set of standard “outputs” , including RDF, OWL, TTL, knowledge graphs, and others.

The platform is already reducing ontology development timelines from months to weeks, sometimes days. The goal is to compress this further, ideally hours to a couple of days.