Biography
Inesh Hettiarachchi is an independent AI researcher and technology leader whose work focuses on temporal intelligence, large language models, and trustworthy enterprise AI. His research explores how AI systems can account for evolving information, preserve historical context, and support transparent, accountable decision-making.
His OrgVerse research proposes a temporal and multidimensional framework for enterprise data integration and AI-assisted decision support, incorporating human oversight and governance. His broader research interests include generative AI agents, temporal data representation, federated learning, and domain adaptation of language models.
With over 25 years of software engineering and enterprise systems experience, Inesh brings a practical perspective to AI research, connecting emerging methods with real-world challenges.
Abstract
Title: OrgVerse: Time-Aware Enterprise AI for Trustworthy Decision-Making
Abstract: Enterprise AI systems must reason over changing records, policies, and organizational relationships. Without temporal context, they can mix historical and current information, producing misleading conclusions. This presentation introduces OrgVerse, my proposed framework for reconstructing organizational state over time to support traceable decision-making. Through a transaction-investigation scenario, I will explore temporal knowledge representation, evidence provenance, access controls, and human oversight. The session will also discuss how to evaluate historical accuracy and information leakage, distinguishing proposed capabilities from validated findings.
