DATA-DRIVEN CHARACTERIZATION OF TECHNICAL DEBT IN A COMPLEX INFORMATION SYSTEM

Presented herein are techniques that provide a holistic and integrated abstraction among different categories of technical debt (TD) in a complex software system, as well as among different TD-related data sources such as logs, traces, telemetry, and metrics. The techniques presented herein allow for accelerated, automated, and evolutionary TD management in a complex software development life cycle (SDLC). The techniques learn the context throughout the SDLC pipeline and turn this context into actionable insights for use in repaying the technical debt at the earliest stages of the development process. The techniques presented herein provide an automated and low cost mechanism that may reduce debt within a company