Mathematical Modelling of the Impact of Developer Experience Metrics on the Duration of the Release Cycle in Full-Stack Projects

Modern DevOps practices are increasingly emphasizing Developer Experience metrics as critical factors influencing the efficacy of software delivery. However, despite significant advancements in this direction, the challenge of constructing a causal model that elucidates the impact of Developer Experience metrics on the speed of releases remains unresolved. This study aims to formalize and empirically evaluate the causal relationship between various indicators of development experience (such as cognitive load and technical frustration) and the speed of the release cycle within DevOps environments. A simulation model was developed for this analysis, embodying a prototypical CI/CD infrastructure characterized by either a microservices or serverless architecture within a cloud framework. Bayesian Causal Modeling was employed, facilitating the exploration of how Developer Experience variables influence the time parameters of the release cycle. The model encompasses seven variables, comprising five key Developer Experience metrics alongside two control parameters (architectural and organizational levels). Within the simulation framework, data was generated based on the principles of Web Application Architecture, incorporating contemporary methodologies of Full Stack Development and RESTful API interfaces. The findings reveal a statistically significant causal relationship between cognitive load and technical frustration and an increase in the duration of the deployment cycle. For instance, the probability of exceeding a release duration of 14 days increases from 21% to 83% in case of slow CI/CD integration, and accordingly from 18% to 72% when utilizing low-quality tools. This underscores the feasibility of integrating Developer Experience metrics into Engineering Decision Support Systems within DevOps processes. The results obtained from this study can be leveraged to enhance DevOps monitoring strategies, prioritize interventions within development workflows, and further advance analytical platforms aimed at managing the quality of the developer experience.