Methodological Approaches to Assessing Productivity and Resource Planning in Hybrid Human-AI Software Engineering Teams

The article is dedicated to the analysis of methodological approaches to assessing productivity and resource planning in hybrid human–AI software engineering teams. The relevance of the study is determined by the rapid integration of generative AI tools into development processes and the growing mismatch between traditional productivity metrics and emerging socio-technical realities. Scientific novelty lies in the analytical reinterpretation of productivity as a dynamic balance between generative speed, stabilization effort, architectural coherence, and shared understanding, rather than as a single output indicator. The work describes how AI reshapes coordination rhythms, redistributes cognitive and organizational effort, and alters the temporal structure of planning and risk manifestation. Special attention is paid to the delayed effects of AI-driven acceleration, including architectural drift, erosion of contextual knowledge, and the growing invisibility of preventive work. The study sets itself the goal of identifying methodological principles suitable for evaluating productivity and planning capacity in hybrid teams. To achieve this goal, analytical synthesis, comparative analysis, and source-based conceptual modeling are applied. The conclusion outlines implications for measurement frameworks and governance practices. The article will be useful for researchers, software engineering managers, and practitioners involved in AI-augmented development environments.