Organizational Coupling as a Leading Indicator of Microservice Architecture Degradation

Publication at a glance

Paper: Organizational Coupling as a Leading Indicator of Microservice Architecture Degradation
Authors: Nariman Mani, Jose Sosa, Xiaozhou Li, Tomas Cerny
Published in: 23rd IEEE International Conference on Software Architecture (ICSA 2026)
Persistent record: ICSA 2026
Google Scholar: View the publication record

The research problem

Architecture degradation is often detected after structural symptoms become expensive: shared changes, tangled dependencies, coordination delays, and blurred ownership. The organization may reveal the drift earlier when developers increasingly span service boundaries.

The paper's central contribution

This work positions organizational coupling as a leading indicator. The hypothesis is not that collaboration is bad; rather, sustained patterns of cross-service work can signal that supposedly independent services require hidden coordination and that implementation boundaries no longer match operating boundaries.

This distinction matters for both researchers and practitioners: the contribution is a way to reason about the complete system and its decision boundaries, rather than a claim that adding one AI model solves the underlying engineering problem.

How the approach works

Version-control histories can derive evolving developer-to-service relationships.
Time-aware coupling trends help separate ordinary collaboration from persistent boundary pressure.
Comparing organizational signals with architectural change supports earlier, evidence-based governance.

Together, these elements make the approach easier to study, reproduce, and extend. They also provide clear terms for literature searches and comparisons with related work.

Why this publication matters

The work matters to research on architecture erosion, technical debt, Conway’s law, service ownership, developer networks, and socio-technical systems. Architecture teams can use the signal alongside dependency and change-coupling analysis to decide where a deeper review is justified.

For practitioners, the larger lesson is to measure the system behavior that matters—not only a model-level score. Reliability, privacy, change over time, human review, and downstream consequences belong in the evaluation.

Research questions this work can support

This publication is relevant when investigating questions such as:

How should the core problem be represented so an AI-assisted system retains the right context?
Which technical and organizational signals provide early, actionable evidence?
How can automation improve adaptability without concealing failures or weakening safeguards?
What evaluation design captures effectiveness, safety, privacy, and long-term change?
Where should a system defer to a developer, architect, coach, clinician, or participant?

Scope and responsible interpretation

A leading indicator is not proof of degradation. Cross-functional platform work, incident response, mentoring, or planned migration can create healthy cross-service activity. Interpretation requires organizational context, multiple measures, and safeguards against using developer analytics for surveillance.

These boundaries are opportunities for replication, comparison, and follow-on research. Readers should consult the paper itself for its precise method, datasets, experimental setup, and reported results.

How to cite this paper

Mani, N., Sosa, J., Li, X., & Cerny, T. (2026). Organizational Coupling as a Leading Indicator of Microservice Architecture Degradation. In IEEE International Conference on Software Architecture (ICSA 2026).

Use the ICSA 2026 publication page to verify the latest bibliographic metadata before submission. You can also find the work through its Google Scholar record. If your research builds on the architecture, method, evaluation, or research framing described here, please cite the original publication rather than this explanatory post.

Related research on this weblog

Browse the complete Nariman Mani publications list.
Explore GraphQL-aware healing with multi-signal learning.
For adjacent work, read adaptive test healing with LLMs and reinforcement learning and privacy-by-design adaptive group assignment.