arXiv subjects

Damian Andrew Tamburri

Publications and source records attributed to Damian Andrew Tamburri.

At least 19 recordsLinked to original sources

arXiv2026

Actionable Meta-Requirements for AI Robustness Auditing

AI-enabled systems must remain dependable when the conditions under which they operate change (e.g., due to input perturbations, distribution shifts, or changes in external dependencies). This makes robustness auditing of such systems important for judging whether the available evidence is sufficient to support a defined robustness conclusion. However, designing an AI robustness auditing framework is challenging. A key unresolved issue is determining the requirements such a framework should support; these requirements have not yet been consolidated. We address this gap through meta-requirements: higher-order requirements specifying what a class of auditing frameworks should support. In design science research, literature and domain experts are common sources of meta-requirements; accordingly, we derive the initial set from literature and validate and refine it with practitioners. Our systematic literature review retained 74 of 2,890 records and derived 43 meta-requirements across 16 categories. We evaluated them with 128 practitioners, obtaining 5,504 ratings and 586 substantive comments. Overall, 81.20% of ratings indicated agreement or strong agreement. In particular, 41 of the 43 literature-derived meta-requirements were classified as strongly supported. Practitioner feedback refined the requirements and identified an additional requirement concerning external components and supply chain dependencies. The final set comprises 44 meta-requirements across 16 categories, of which 43 received quantitative validation. These meta-requirements can help organizations implement AI robustness auditing frameworks, assess existing frameworks for completeness, and refine them.

cs.AI↗

arXiv2026

Chaos Engineering in the Wild: Findings from GitHub

Chaos Engineering (CE) enhances system resilience by deliberately injecting faults to uncover weaknesses before they cause outages. Despite the availability of many CE tools, little is known about how they are adopted and maintained in open-source software (OSS) projects. This study empirically characterizes the adoption, evolution, and practical use of CE tools in OSS projects, examining who adopts them, how they are maintained, and which fault types they target. We conducted a large-scale mining study of GitHub repositories associated with ten widely used CE tools. Starting from 5,410 candidate repositories, we systematically filtered and manually validated 1,275 records and analyzed repository metadata, commit histories, and documentation. We found that adoption is concentrated around a few tools, with Toxiproxy, Chaos Mesh, and Chaos Monkey accounting for 68.86% of the validated repositories. In terms of adopter context, development is the predominant repository purpose (56.55%), while industry represents the largest ownership category (34.82%), closely followed by personal repositories (33.49%). At the project level, activity varies substantially, with 51.42% of repositories having at most 50 commits. In terms of fault coverage, network faults (44.85%) and instance termination (29.96%) together account for 74.81% of the 2,410 observed fault instances, whereas application-level faults account for only 2.57%. Taken together, these findings suggest that practitioners should consider repository activity and fault coverage when selecting CE tools, while researchers should investigate the factors behind concentrated adoption and limited application-level experimentation.

cs.SE↗

arXiv2026

A Taxonomy of Runtime Faults in Model Context Protocol Servers

MCP (Model Context Protocol) enables LLMs (Large Language Models) to interact with external tools and data sources via a standardized protocol. Its rapid adoption in tool-augmented Artificial Intelligence (AI) workflows has introduced new reliability challenges, such as configuration parameters that are accepted but not enforced at runtime, leading to unintended default behavior, whose runtime fault characteristics remain empirically unexamined. We present the first empirical taxonomy of runtime faults in MCP servers. We manually analyzed 833 confirmed runtime fault threads identified through a repository-mining pipeline covering 473 actively maintained MCP server GitHub repositories and derived a taxonomy using a bottom-up open coding procedure. The taxonomy comprises 11 top-level categories and 27 subcategories (73 leaf fault types), covering recurrent failures across protocol interactions, tool invocations, schema enforcement, state management, model-provider integration, security validation, and timeouts or explicit cancellations of in-progress operations. To assess the taxonomy's external validity, we surveyed 55 MCP server developers. Respondents reported experiencing an average of 20 of the 27 fault subcategories, and no category remained unobserved. These results indicate that the taxonomy reflects widely observed runtime failures in MCP-based systems and shall assist AI software maintenance and evolution in the future.

cs.SE↗

arXiv2026

"Show Me You Comply... Without Showing Me Anything": Zero-Knowledge Software Auditing for AI-Enabled Systems

Classical software verification and validation techniques, such as procedural audits, formal methods, or model documentation, are the traditional mechanisms used to achieve the verifiable accountability now required by regulations like the EU AI Act. These methods are either expensive or heavily manual, and ill-suited for the opaque, "black box" nature of most Artificial Intelligence (AI) models. A conflict arises: high auditability and verifiability are required by law, but such transparency conflicts with the need to protect the assets being audited (e.g., confidential data and proprietary models). This paper introduces ZKMLOps, an \ac{MLOps} verification framework that operationalizes Zero-Knowledge Proofs (ZKPs) within Machine-Learning Operations lifecycles; a ZKP allows a prover to convince a verifier that a statement is true without revealing any information about the statement itself. By integrating ZKP with established software engineering patterns, ZKMLOps provides a modular and repeatable process for generating verifiable cryptographic evidence-proofs of well-defined computational statements about the audited model and its inputs-that auditors can use as input to a regulatory compliance determination. We evaluate the framework along two dimensions. First, framework viability: orchestration overhead is bounded and stable across architecturally heterogeneous ZKP backends and models of increasing size. Second, cost-versus-assurance trade-offs: the audit-on-demand setting is the regime in which full zero-knowledge auditing is the appropriate tool, where it provides confidentiality and integrity guarantees that lighter-weight alternatives cannot match.

cs.SE↗

arXiv2026

Privacy Engineering: A Systematic Literature Review

Privacy obligations under GDPR increasingly shape software engineering. We synthesize 90 studies from 2018 to 2025 using a systematic review with thematic synthesis to chart privacy engineering. Thirteen dimensions form two recurrent cores: Privacy Enhancing Technologies (PETs) with Privacy Metrics (PM) and Verification and Testing (VT) and Governance and Accountability (GA) with Transparency and Communication (TC) and Organizational Measures (OM). Modeling and Specification (MS) mediates between the cores. Lifecycle mapping shows concentrations at requirements and design (MS, GA), at implementation and verification (PETs, VT, PM, TC) and at operation and decommissioning (GA, OM, Data Subject Rights Management (DSRM), Incident Response and Management (IRM), Lifelong Management (LM)). Handoffs link models to rules and tests, mechanisms to metrics and deployments such as enclaves and ledgers to governance records. Domains reweight but do not alter structure: healthcare weights GA with VT and PETs, IoT and edge weight PETs with VT and PM at device and edge, web measurement weights TC with VT, AI and ML weight PETs with PM. IRM, LM and Data Minimization and Purpose Limitation (DMPL) are less often primary foci, signaling priorities for future work. The results provide a practical map and a replication-ready scaffold for assessment and updates.

cs.SE↗

arXiv2026

How are MLOps Frameworks Used in Open Source Projects? An Empirical Characterization

Machine Learning (ML) Operations (MLOps) frameworks have been conceived to support developers and AI engineers in managing the lifecycle of their ML models. While such frameworks provide a wide range of features, developers may leverage only a subset of them, while missing some highly desired features. This paper investigates the practical use and desired feature enhancements of eight popular open-source MLOps frameworks. Specifically, we analyze their usage by dependent projects on GitHub, examining how they invoke the frameworks' APIs and commands. Then, we qualitatively analyze feature requests and enhancements mined from the frameworks' issue trackers, relating these desired improvements to the previously identified usage features. Results indicate that MLOps frameworks are rarely used out-of-the-box and are infrequently integrated into GitHub Workflows, but rather, developers use their APIs to implement custom functionality in their projects. Used features concern core ML phases and whole infrastructure governance, sometimes leveraging multiple frameworks with complementary features. The mapping with feature requests highlights that users mainly ask for enhancements to core features of the frameworks, but also better API exposure and CI/CD integration.

cs.SE↗

arXiv2025

IaC Generation with LLMs: An Error Taxonomy and A Study on Configuration Knowledge Injection

Large Language Models (LLMs) currently exhibit low success rates in generating correct and intent-aligned Infrastructure as Code (IaC). This research investigated methods to improve LLM-based IaC generation, specifically for Terraform, by systematically injecting structured configuration knowledge. To facilitate this, an existing IaC-Eval benchmark was significantly enhanced with cloud emulation and automated error analysis. Additionally, a novel error taxonomy for LLM-assisted IaC code generation was developed. A series of knowledge injection techniques was implemented and evaluated, progressing from Naive Retrieval-Augmented Generation (RAG) to more sophisticated Graph RAG approaches. These included semantic enrichment of graph components and modeling inter-resource dependencies. Experimental results demonstrated that while baseline LLM performance was poor (27.1% overall success), injecting structured configuration knowledge increased technical validation success to 75.3% and overall success to 62.6%. Despite these gains in technical correctness, intent alignment plateaued, revealing a "Correctness-Congruence Gap" where LLMs can become proficient "coders" but remain limited "architects" in fulfilling nuanced user intent.

cs.AI↗

arXiv2025

"Let it be Chaos in the Plumbing!" Usage and Efficacy of Chaos Engineering in DevOps Pipelines

Chaos Engineering (CE) has emerged as a proactive method to improve the resilience of modern distributed systems, particularly within DevOps environments. Originally pioneered by Netflix, CE simulates real-world failures to expose weaknesses before they impact production. In this paper, we present a systematic gray literature review that investigates how industry practitioners have adopted and adapted CE principles over recent years. Analyzing 50 sources published between 2019 and early 2024, we developed a comprehensive classification framework that extends the foundational CE principles into ten distinct concepts. Our study reveals that while the core tenets of CE remain influential, practitioners increasingly emphasize controlled experimentation, automation, and risk mitigation strategies to align with the demands of agile and continuously evolving DevOps pipelines. Our results enhance the understanding of how CE is intended and implemented in practice, and offer guidance for future research and industrial applications aimed at improving system robustness in dynamic production environments.

cs.SE↗

arXiv2025

Chaos Engineering: A Multi-Vocal Literature Review

Organizations, particularly medium and large enterprises, typically rely heavily on complex, distributed systems to deliver critical services and products. However, the growing complexity of these systems poses challenges in ensuring service availability, performance, and reliability. Traditional resilience testing methods often fail to capture the intricate interactions and failure modes of modern systems. Chaos Engineering addresses these challenges by proactively testing how systems in production behave under turbulent conditions, allowing developers to uncover and resolve potential issues before they escalate into outages. Though chaos engineering has received growing attention from researchers and practitioners alike, we observed a lack of reviews that synthesize insights from both academic and grey literature. Hence, we conducted a Multivocal Literature Review (MLR) on chaos engineering to address this research gap by systematically analyzing 96 academic and grey literature sources published between January 2016 and April 2024. We first used the chosen sources to derive a unified definition of chaos engineering and to identify key functionalities, components, and adoption drivers. We also developed a taxonomy for chaos engineering platforms and compared the relevant tools using it. Finally, we analyzed the current state of chaos engineering research and identified several open research issues.

cs.SE↗

arXiv2025

Recursive Self-Similarity in Deep Weight Spaces of Neural Architectures: A Fractal and Coarse Geometry Perspective

This paper conceptualizes the Deep Weight Spaces (DWS) of neural architectures as hierarchical, fractal-like, coarse geometric structures observable at discrete integer scales through recursive dilation. We introduce a coarse group action termed the fractal transformation, $T_{r_k} $, acting under the symmetry group $G = (\mathbb{Z}, +) $, to analyze neural parameter matrices or tensors, by segmenting the underlying discrete grid $Ω$ into $N(r_k)$ fractals across varying observation scales $ r_k $. This perspective adopts a box count technique, commonly used to assess the hierarchical and scale-related geometry of physical structures, which has been extensively formalized under the topic of fractal geometry. We assess the structural complexity of neural layers by estimating the Hausdorff-Besicovitch dimension of their layers and evaluating a degree of self-similarity. The fractal transformation features key algebraic properties such as linearity, identity, and asymptotic invertibility, which is a signature of coarse structures. We show that the coarse group action exhibits a set of symmetries such as Discrete Scale Invariance (DSI) under recursive dilation, strong invariance followed by weak equivariance to permutations, alongside respecting the scaling equivariance of activation functions, defined by the intertwiner group relations. Our framework targets large-scale structural properties of DWS, deliberately overlooking minor inconsistencies to focus on significant geometric characteristics of neural networks. Experiments on CIFAR-10 using ResNet-18, VGG-16, and a custom CNN validate our approach, demonstrating effective fractal segmentation and structural analysis.

cs.NE↗

arXiv2024

Data Mesh: a Systematic Gray Literature Review

Data mesh is an emerging domain-driven decentralized data architecture that aims to minimize or avoid operational bottlenecks associated with centralized, monolithic data architectures in enterprises. The topic has picked the practitioners' interest, and there is considerable gray literature on it. At the same time, we observe a lack of academic attempts at defining and building upon the concept. Hence, in this article, we aim to start from the foundations and characterize the data mesh architecture regarding its design principles, architectural components, capabilities, and organizational roles. We systematically collected, analyzed, and synthesized 114 industrial gray literature articles. The review provides insights into practitioners' perspectives on the four key principles of data mesh: data as a product, domain ownership of data, self-serve data platform, and federated computational governance. Moreover, due to the comparability of data mesh and SOA (service-oriented architecture), we mapped the findings from the gray literature into the reference architectures from the SOA academic literature to create the reference architectures for describing three key dimensions of data mesh: organization of capabilities and roles, development, and runtime. Finally, we discuss open research issues in data mesh, partially based on the findings from the gray literature.

cs.SE↗

arXiv2024

Architectural Design Decisions for Self-Serve Data Platforms in Data Meshes

Data mesh is an emerging decentralized approach to managing and generating value from analytical enterprise data at scale. It shifts the ownership of the data to the business domains closest to the data, promotes sharing and managing data as autonomous products, and uses a federated and automated data governance model. The data mesh relies on a managed data platform that offers services to domain and governance teams to build, share, and manage data products efficiently. However, designing and implementing a self-serve data platform is challenging, and the platform engineers and architects must understand and choose the appropriate design options to ensure the platform will enhance the experience of domain and governance teams. For these reasons, this paper proposes a catalog of architectural design decisions and their corresponding decision options by systematically reviewing 43 industrial gray literature articles on self-serve data platforms in data mesh. Moreover, we used semi-structured interviews with six data engineering experts with data mesh experience to validate, refine, and extend the findings from the literature. Such a catalog of design decisions and options drawn from the state of practice shall aid practitioners in building data meshes while providing a baseline for further research on data mesh architectures.

cs.SE↗

arXiv2023

Architecture Smells vs. Concurrency Bugs: an Exploratory Study and Negative Results

Technical debt occurs in many different forms across software artifacts. One such form is connected to software architectures where debt emerges in the form of structural anti-patterns across architecture elements, namely, architecture smells. As defined in the literature, ``Architecture smells are recurrent architectural decisions that negatively impact internal system quality", thus increasing technical debt. In this paper, we aim at exploring whether there exist manifestations of architectural technical debt beyond decreased code or architectural quality, namely, whether there is a relation between architecture smells (which primarily reflect structural characteristics) and the occurrence of concurrency bugs (which primarily manifest at runtime). We study 125 releases of 5 large data-intensive software systems to reveal that (1) several architecture smells may in fact indicate the presence of concurrency problems likely to manifest at runtime but (2) smells are not correlated with concurrency in general -- rather, for specific concurrency bugs they must be combined with an accompanying articulation of specific project characteristics such as project distribution. As an example, a cyclic dependency could be present in the code, but the specific execution-flow could be never executed at runtime.

cs.SE↗

arXiv2022

Microservice Architecture Practices and Experience: a Focused Look on Docker Configuration Files

Cloud applications are more and more microservice-oriented, but a concrete charting of the microservices architecture landscape -- namely, the space of technical options available for microservice software architects in their decision-making -- is still very much lacking, thereby limiting the ability of software architects to properly evaluate their architectural decisions with sound experiential devices and/or practical design principles. On the one hand, Microservices are fine-grained, loosely coupled services that communicate through lightweight protocols. On the other hand, each microservice can use a different software stack, be deployed and scaled independently or even executed in different containers, which provide isolation and a wide-range of configuration options but also offer unforeseeable architectural interactions and under-explored architecture smells, with such experience captured mainly in software repositories where such solutions are cycled. This paper adopts a mining software repositories (MSR) approach to capture the practice within the microservice architecture landscape, by eliciting and analysing Docker configuration files, being Docker the leading technical device to design for, and implement modern microservices. Our analysis of Docker-based microservices gives an interesting summary of the current state of microservices practice and experience. Conversely, observing that all our datapoints have their own shape and characteristics, we conclude that further comparative assessment with industrial systems is needed to better address the recurring positive principles and patterns around microservices.

cs.SE↗

arXiv2022

Blockchain-Oriented Services Computing in Action: Insights from a User Study

Blockchain architectures promise disruptive innovation but factually they pose many architectural restrictions to classical service-based applications and show considerable design, implementation, and operations overhead. Furthermore, the relation between such overheads and user benefits is not clear yet. To shed light on the aforementioned relations, a service-based blockchain architecture was designed and deployed as part of a field study in real-life experimentation. An observational approach was then performed to elaborate on the technology-acceptance of the service-based blockchain architecture in question. Evidence shows that the resulting architecture is, in principle, not different than other less complex equivalents; furthermore, the architectural limitations posed by the blockchain-oriented design demand a significant additional effort to be put onto even the simplest of functionalities. We conclude that further research shall be invested in clarifying further the design principles we learned as part of this study as well as any trade-offs posed by blockchain-oriented service design and operation.

cs.SE↗

arXiv2022

A Declarative Modelling Framework for the Deployment and Management of Blockchain Applications

The deployment and management of Blockchain applications require non-trivial efforts given the unique characteristics of their infrastructure (i.e., immutability) and the complexity of the software systems being executed. The operation of Blockchain applications is still based on ad-hoc solutions that are error-prone, difficult to maintain and evolve, and do not manage their interactions with other infrastructures (e.g., a Cloud backend). This paper proposes KATENA, a framework for the deployment and management of Blockchain applications. In particular, it focuses on applications that are compatible with Ethereum, a popular general-purpose Blockchain technology. KATENA provides i) a metamodel for defining Blockchain applications, ii) a set of processes to automate the deployment and management of defined models, and iii) an implementation of the approach based on TOSCA, a standard language for Infrastructure-as-Code, and xOpera, a TOSCA-compatible orchestrator. To evaluate the approach, we applied KATENA to model and deploy three real-world Blockchain applications, and showed that our solution reduces the amount of code required for their operations up to $82.7\%$.

cs.SE↗

arXiv2021

DataOps for Societal Intelligence: a Data Pipeline for Labor Market Skills Extraction and Matching

Big Data analytics supported by AI algorithms can support skills localization and retrieval in the context of a labor market intelligence problem. We formulate and solve this problem through specific DataOps models, blending data sources from administrative and technical partners in several countries into cooperation, creating shared knowledge to support policy and decision-making. We then focus on the critical task of skills extraction from resumes and vacancies featuring state-of-the-art machine learning models. We showcase preliminary results with applied machine learning on real data from the employment agencies of the Netherlands and the Flemish region in Belgium. The final goal is to match these skills to standard ontologies of skills, jobs and occupations.

cs.AI↗

arXiv2020

Blockchain and Cryptocurrencies: a Classification and Comparison of Architecture Drivers

Blockchain is a decentralized transaction and data management solution, the technological leap behind the success of Bitcoin and other cryptocurrencies. As the variety of existing blockchains and distributed ledgers continues to increase, adopters should focus on selecting the solution that best fits their needs and the requirements of their decentralized applications, rather than developing yet another blockchain from scratch. In this paper we present a conceptual framework to aid software architects, developers, and decision makers to adopt the right blockchain technology. The framework exposes the interrelation between technological decisions and architectural features, capturing the knowledge from existing academic literature, industrial products, technical forums/blogs, and experts' feedback. We empirically show the applicability of our framework by dissecting the platforms behind Bitcoin and other top 10 cryptocurrencies, aided by a focus group with researchers and industry practitioners. Then, we leverage the framework together with key notions of the Architectural Tradeoff Analysis Method (ATAM) to analyze four real-world blockchain case studies from industry and academia. Results shown that applying our framework leads to a deeper understanding of the architectural tradeoffs, allowing to assess technologies more objectively and select the one that best fit developers needs, ultimately cutting costs, reducing time-to-market and accelerating return on investment.

cs.CR↗