Demand-Driven Context: A Methodology for Building Enterprise Knowledge Bases Through Agent Failure

ArXi:2603.14057v1 Announce Type: new Large language model agents nstrate expert-level reasoning, yet consistently fail on enterprise-specific tasks due to missing domain knowledge -- terminology, operational procedures, system interdependencies, and institutional decisions that exist largely as tribal knowledge. Current approaches fall into two categories: top-down knowledge engineering, which documents domain knowledge before agents use it, and bottom-up automation, where agents learn from task experience.