Modeling the balance between cognitive focus and team effectiveness in intelligent notification filtering systems within Agile teams

Abstract

The purpose of the study is to develop a mathematical model for adaptive notification filtering that makes it possible to optimize the balance between an individual developer’s cognitive focus and collective efficiency (Velocity) within a dynamic Agile environment. Method. The research methodology is based on an integrated combination of graph theory to model dynamic logical dependencies among tasks performed by members of an Agile team, and queueing theory to analyze incoming notification streams. To formalize individual cognitive losses, the study employs mathematical modeling of nonlinear exponential relationships that describe the process of concentration recovery («cognitive viscosity») as a function of uninterrupted work duration. Team-level losses are assessed by computing the expected value of idle time through stochastic analysis of specialists’ backlog states and by determining node centrality metrics within the project dependency graph. Optimization of notification delivery timing is achieved by solving a global objective maximization problem using adaptive weighting coefficients that vary in accordance with the project phase. Results. The study resulted is a mathematical model and an intelligent notification-filtering algorithm that enable quantitative balancing between an individual developer’s productivity and overall team velocity. The research formalizes an autonomy buffer criterion that allows the system to anticipate critical moments when colleagues’ parallel tasks are nearing exhaustion and to proactively prevent blockage of the project’s critical path. It is demonstrated that the use of a dynamic sprint-progress coefficient ensures a flexible transition from an aggressive cognitive-focus protection strategy at the beginning of an iteration to the prioritization of immediate communication during the release phase. Situational analysis confirms that the proposed approach makes it possible to transform the subjective importance of messages into measurable economic and temporal metrics (Cost of Delay, Lead Time), minimizing losses caused by interruptions to the flow state without compromising team synchronization. The scientific novelty of the study lies in the development of a comprehensive mathematical model for adaptive notification filtering that, for the first time, integrates the «cognitive viscosity» of an individual developer with a dynamic graph of team dependencies within the Agile cycle. Unlike existing approaches that treat the specialist as an isolated unit, the proposed model formalizes the nonlinear relationship between concentration recovery time and the depth of immersion in the «flow state», and introduces an adaptive sprint-progress coefficient to achieve Paretoefficient balancing between the protection of individual cognitive resources and the minimization of team-level losses. This makes it possible to quantitatively justify the transition from static filtering regimes to context-aware communication management, in which message delivery priorities dynamically adjust according to the project phase and the probability of blocking the critical path. The practical significance of the obtained results lies in establishing an algorithmic and methodological foundation for the development of intelligent notification filtering systems capable of dynamically adapting to the phases of an IT project’s life cycle. Implementing the proposed models in corporate messengers and project management systems will make it possible to automate the calculation of an «autonomy buffer» for each developer and to apply an intelligent message «batching» mechanism. This will enable the minimization of team losses (Cost of Delay) and the optimization of overall development speed (Lead Time) by eliminating cascading idle time caused by communication delays along critical paths. The proposed approach makes it possible to transform the subjective importance of information flows into a measurable quantity, ensuring a balance between protecting an individual’s «flow state» and achieving the collective goals of an Agile team.