arXiv subjects

Mirko Perkusich

Publications and source records attributed to Mirko Perkusich.

8 recordsLinked to original sources

arXiv2026

From Textual Requirements to Microservice Architectures - A Comprehensive Evaluation of LLM-Based Design Synthesis

Microservice architectures have become dominant for modernizing monolithic systems, yet identifying appropriate services remains challenging and largely manual. Existing decomposition approaches are predominantly code-centric, limiting applicability in early design stages where only textual requirements are available. Despite advances in Large Language Models (LLMs), limited empirical evidence exists on their ability to synthesize complete microservice architectures from natural-language requirements, including service definitions and inter-service interactions. This study investigates whether an LLM can bridge requirements engineering and architectural design, generating architectures solely from textual requirements and evaluating structural agreement and perceived quality of results. We conduct a mixed-method study using OpenAI o3 under zero-shot (ZS) and few-shot (FS) prompting across two systems (Bookstore, PetClinic), one execution per system/condition. Architectures are evaluated through (i) comparison with reference architectures using precision, recall, and F1-score for service identification and communication recovery, and (ii) a blinded expert assessment of correctness, completeness, modularity, and plausibility, plus open feedback synthesis. OpenAI o3 identifies services with higher agreement under FS prompting (F1 = 0.79 for ZS versus = 0.97 for FS). Communication recovery is more challenging: ZS produces dense architectures with high recall but low precision (F1 = 0.61), while FS improves agreement, reaching F1 = 0.82 and reducing unsupported dependencies. Expert evaluation corroborates these results, with FS architectures perceived as more modular, coherent, and plausible than ZS outputs. OpenAI o3 shows potential for requirements-driven synthesis when guided by exemplar prompting. Results are model- and context-specific from two small systems, not model-independent proof.

cs.SE↗

arXiv2026

Structural Validation of LLM-Generated Microservice Decompositions Using Source-Code Dependencies

Decomposing monolithic systems into microservices is a key activity in software modernization. Although Large Language Models (LLMs) can generate semantically plausible decompositions from textual requirements, it remains unclear whether these proposals preserve the structural dependencies implemented in the source code. This paper evaluates the structural adherence of microservice decompositions generated by OpenAI o3 for the PetClinic and Bookstore systems. We propose an automated validation pipeline based on static dependency analysis and compare zero-shot and few-shot prompting using dependency preservation (TPD) and dependency violation (TVD) metrics. A robustness analysis was conducted to control for differences in class-to-service mapping coverage. After normalization, both prompting strategies produced equivalent structural adherence, achieving TPD values of 68.0% (PetClinic) and 83.3% (Bookstore). The findings demonstrate that structural evaluations of LLM-generated decompositions should explicitly control for mapping coverage, as apparent differences between prompting strategies may otherwise reflect methodological bias rather than genuine architectural quality.

cs.SE↗

arXiv2026

AI-Conducted Interviews in Empirical Software Engineering: An Experience Report

Semi-structured interviews are widely used in empirical software engineering (ESE), but they are resource-intensive and difficult to coordinate across schedules, locations, and natural languages. This experience report examines a customized MyGPT used to conduct short, self-administered interviews in two ESE studies: one on refactoring practices and another on generative AI in Scrum-related activities. Participants accessed the interviewer through shared links, used voice interaction, selected a preferred natural language, and completed the interview without a researcher present. The AI followed a predefined protocol and generated a structured synthesis that participants voluntarily submitted; these artifacts were not treated as verbatim transcripts. We analyzed 66 submissions and questionnaire responses, and audited artifact format, language, length, and protocol consistency. Of the submitted artifacts, 92.4% followed the expected synthesis format, 65 were predominantly in Portuguese and one in English, and two conflicted with the reported protocol. Participants generally rated the experience positively: 90.9% reported a positive overall experience and comfort, 95.5% considered the questions clear, 97.0% rated the pace positively, and 89.4% would participate again. Reported limitations included generic questions, limited sensitivity to answers, insufficient depth, privacy concerns, and missed human interaction. The findings support the operational viability and acceptability of this workflow among analyzed respondents, but do not establish completion rates, time savings, summary fidelity, or equivalence to human-conducted interviews. AI interviewers should therefore be treated as a complementary option for short, focused, low-risk studies, with protocol design, privacy guidance, artifact validation, and human oversight.

cs.SE↗

arXiv2026

Comparing Large Language Models on Scrum Certification-Style Questions: Accuracy, Stability, and Error Patterns

Large Language Models (LLMs) are increasingly used in exam- and certification-style question answering tasks, where their ability to retrieve, interpret, and apply domain-specific knowledge can be systematically assessed. In Software Engineering, such settings are particularly relevant when questions depend on strict adherence to normative definitions, roles, artifacts, and rules. This paper evaluates the performance of three contemporary LLMs, \textit{GPT-5 mini}, \textit{Gemini 3 Flash}, and \textit{DeepSeek Chat 3.2}, in answering 993 Scrum certification-style questions aligned with the Professional Scrum Master I (PSM I) assessment format. We evaluated the models under three prompting strategies (\textit{zero-shot}, \textit{chain-of-thought}, and \textit{source-grounded}), with repeated executions to assess intra-model stability. We also analyzed performance across Scrum topics and question formats, complemented by a qualitative analysis of recurring error patterns in incorrect answers. Results revealed clear differences among models, with Gemini 3 Flash achieving the highest accuracy, followed by GPT-5 mini and DeepSeek Chat 3.2, while intra-model variability remained low across all conditions. By question format, the models achieved the highest accuracy on single-answer multiple-choice items, whereas multi-select and True/False questions were more error-prone. By topic, performance was more consistent in normatively explicit areas such as Artifacts, Empiricism, and Product Value, but more fragile in Scrum Values, Self-Managing Teams, and Stakeholders \& Customers. The qualitative analysis showed that errors were systematic rather than random, involving overgeneralization, restrictive wording, compound distractors, and conflicts between common market interpretations and strict Scrum definitions.

cs.SE↗

arXiv2026

Prompting GPT-5 on Scrum Certification Questions: An Empirical Accuracy Study

Large Language Models (LLMs) are increasingly used in Agile Software Development for documentation, coaching, and training. As practitioners adopt these tools to prepare for certifications such as Professional Scrum Master (PSM), a key question is whether LLMs can reliably reason about Scrum, a framework with normative, well-defined rules described in the Scrum Guide (2020). This paper examines how different prompt techniques affect the factual accuracy of LLM responses to Scrum certification-style questions. A dataset of 993 validated PSM-aligned questions was answered by GPT-5 using three techniques: zero-shot, chain-of-thought, and with-source citation. All prompts achieved certification-level accuracy above 85\%, with the citation-based variant performing best (89.1\%) and yielding the lowest error rate. Correct answers concentrated in well-defined topics, such as \emph{Definition of Done}, Events, and Product Backlog Management, and in single-answer multiple-choice items, while multi-select questions and more interpretive areas, such as Scrum Team and Product Value, were less stable. Among questions where at least one prompt failed (16.2\%), errors clustered into misalignment with the Scrum Guide (28\%), content outside its scope (34\%), and outdated or biased interpretations (38\%). Overall, prompt techniques produced modest but consistent improvements, particularly in reducing misinterpretation and version drift, supporting more reliable use of LLMs in Agile learning and certification preparation.

cs.SE↗

arXiv2026

On the Sensitivity of Firing Rate-Based Federated Spiking Neural Networks to Differential Privacy

Federated Neuromorphic Learning (FNL) enables energy-efficient and privacy-preserving learning on devices without centralizing data. However, real-world deployments require additional privacy mechanisms that can significantly alter training signals. This paper analyzes how Differential Privacy (DP) mechanisms, specifically gradient clipping and noise injection, perturb firing-rate statistics in Spiking Neural Networks (SNNs) and how these perturbations are propagated to rate-based FNL coordination. On a speech recognition task under non-IID settings, ablations across privacy budgets and clipping bounds reveal systematic rate shifts, attenuated aggregation, and ranking instability during client selection. Moreover, we relate these shifts to sparsity and memory indicators. Our findings provide actionable guidance for privacy-preserving FNL, specifically regarding the balance between privacy strength and rate-dependent coordination.

cs.LG↗

arXiv2026

Adoption of Large Language Models in Scrum Management: Insights from Brazilian Practitioners

Scrum is widely adopted in software project management due to its adaptability and collaborative nature. The recent emergence of Large Language Models (LLMs) has created new opportunities to support knowledge-intensive Scrum practices. However, existing research has largely focused on technical activities such as coding and testing, with limited evidence on the use of LLMs in management-related Scrum activities. In this study, we investigate the use of LLMs in Scrum management activities through a survey of 70 Brazilian professionals. Among them, 49 actively use Scrum, and 33 reported using LLM-based assistants in their Scrum practices. The results indicate a high level of proficiency and frequent use of LLMs, with 85% of respondents reporting intermediate or advanced proficiency and 52% using them daily. LLM use concentrates on exploring Scrum practices, with artifacts and events receiving targeted yet uneven support, whereas broader management tasks appear to be adopted more cautiously. The main benefits include increased productivity (78%) and reduced manual effort (75%). However, several critical risks remain, as respondents report 'almost correct' outputs (81%), confidentiality concerns (63%), and hallucinations during use (59%). This work provides one of the first empirical characterizations of LLM use in Scrum management, identifying current practices, quantifying benefits and risks, and outlining directions for responsible adoption and integration in Agile environments.

cs.SE↗

arXiv2026

An Empirical Study of Foundation Models for Variability-Induced Compilation Errors in Configurable C Code

In configurable systems, conditional compilation can hide compilation errors under untested feature combinations. We investigate foundation models for detecting such errors and, in a controlled setting, restoring compilability in configurable C code. Study I evaluates GPT-OSS-20B on 5,000 synthetic snippets generated by ChatGPT-5.2 from 30 curated seeds and exhaustively compiled under all Boolean feature assignments; it also compares TypeChef and evaluates Gemini 3.6 Flash on a stratified sample. GPT-OSS-20B achieved 84.7% micro-precision and 52.1% micro-recall for affected configurations. Coverage depended on reporting style: presence conditions covered 99.4% of failing configurations, whereas explicit enumerations covered 29.5% under a prompt requesting only a minimal justifiable set. GPT-OSS-20B restored compilability for 1,930 of 2,665 faulty snippets (72.4%), while Gemini 3.6 Flash did so for 182 of 190 sampled faulty snippets (95.8%). A paired counterfactual audit found no evidence that an identified label-correlated #define property materially influenced GPT-OSS-20B's predictions. Study II evaluates Codex-GPT5.5 on 100 faulty file-level subjects from five mature configurable systems and reports target-fault-aligned problems in 94 subjects, including four of five historical bugs. Overall, foundation models can support localized detection, explanation, and triage, but should complement compiler-based and variability-aware analyses; compiler acceptance does not establish semantic correctness

cs.SE↗