artificial intelligence

Small and medium-sized enterprises (SMEs) face significant challenges in operationalizing AI, primarily due to constrained resources, limited AI expertise, and the absence of lightweight, production-ready engineering and MLOps support. To address this gap, we propose the first lightweight AI engineering and MLOps blueprint framework specifically designed for SMEs. It integrates domain-customized reference architectures, automated toolchains, and iterative on-site validation mechanisms. Unlike generic enterprise-grade solutions, our blueprint prioritizes low entry barriers, high component reusability, and rapid deployment across the full AI lifecycle—encompassing model development, delivery, and operations. Empirical evaluation across multiple real-world business scenarios demonstrates an average 40% reduction in model delivery time and substantially improved development repeatability. Developer interviews confirm marked reductions in both technical adoption barriers and operational complexity. This work advances the scalable transfer of AI engineering practices from large enterprises to SMEs.

This study addresses the prevailing gap in AI education, which emphasizes model development while neglecting system engineering practices, leaving students ill-equipped to handle real-world challenges such as architectural design, deployment, and monitoring. To bridge this gap, the authors implemented a master’s-level course in which students built a movie recommendation system under realistic constraints, with a focus on integrating AI components into robust software systems, adopting data-driven machine learning practices, and cultivating systems-level thinking. Using a mixed-methods approach—combining analysis of student project artifacts with survey data—the research evaluates learners’ performance in architectural decision-making, integration of heterogeneous models, and adaptation to evolving requirements. Findings reveal common difficulties students encounter in AI system engineering and demonstrate the course’s effectiveness in addressing critical deficiencies in AI engineering education and enhancing systems-aware competencies.

Public-sector deployment of machine learning systems faces dual challenges: technical—such as extreme class imbalance and data drift—and organizational—including bureaucratic data access, unversioned datasets, and absent governance feedback loops. This paper reframes ML pipelines as “civic infrastructure,” integrating LLM-assisted pre-annotation, multi-stage routing classifiers, and controllable synthetic data generation to build an auditable pipeline with data provenance, real-time monitoring, and human-in-the-loop validation. Its core contribution lies in prioritizing institutionalized data engineering—not merely model optimization—to ensure transparency, reproducibility, and accountability. Empirical evaluation on Brazil’s Brasil Participativo platform demonstrates significant improvements in system sustainability and public trust. The framework establishes a transferable engineering paradigm for responsible public AI governance. (149 words)

Medical AI deployment is hindered by insufficient production readiness of machine learning (ML) training pipelines. Method: This paper presents a progressive architectural evolution path—monolithic (chaotic) → modular monolithic → microservices—using SPIRA, a voice-based pre-diagnostic system for respiratory insufficiency, as a case study. It systematically introduces continuous training (CT) and a software-quality-attribute-driven MLOps governance framework tailored to healthcare, integrating modular design, microservice decomposition, and engineered CI/CD pipelines. Contribution/Results: The approach significantly improves pipeline maintainability, fault tolerance, and scalability, enabling stable, iterative evolution of SPIRA. It establishes an “agile ML + robust software engineering” co-design paradigm, delivering a reusable methodology and practical benchmark for engineering medical AI in highly regulated environments.

Manual intervention in ML Ops model deployment hinders responsiveness to model drift and sudden performance degradation. Method: We propose the first multi-armed bandit (MAB)-based reinforcement learning framework for online model evaluation, automatic model selection, and sub-second rollback—integrating UCB, ε-greedy, and Thompson Sampling strategies with real-time performance monitoring and hot-swapping mechanisms, enabling unsupervised, low-latency autonomous decision-making. Results: Experiments on two industrial-scale streaming datasets show that our RL approach improves deployment stability by 23% and reduces human intervention frequency by 91% compared to A/B testing and validation-set-based baselines; rollback latency is compressed to seconds while maintaining or exceeding baseline performance. This work provides the first systematic empirical validation of MAB efficacy in dynamic, production-grade ML Ops model management, extending reinforcement learning from model training into the end-to-end production deployment loop.

This study addresses the degradation of AI model performance in semiconductor manufacturing caused by process variations, equipment aging, and raw material shifts. Leveraging five years of real production line data, the work systematically evaluates multiple MLOps retraining strategies for predictive quality and integrates conformal prediction to deliver statistically valid uncertainty quantification. The authors propose an efficient fixed-interval retraining strategy—updating the model every five lots without hyperparameter tuning—that maintains high prediction accuracy under both abrupt process shifts and gradual equipment degradation while substantially reducing computational overhead. By combining normalized residual control limits with conformal prediction intervals, the approach transitions quality assurance from reactive inspection to proactive, reliable forecasting, offering a robust and practical solution for industrial AI deployment.

This study addresses the lack of systematic comparisons among mainstream MLOps tools, which hinders developers’ ability to select appropriate solutions for their specific needs. For the first time, it presents a multi-dimensional empirical evaluation of MLflow, Metaflow, Apache Airflow, and Kubeflow Pipelines under a unified experimental setup, using two representative tasks: MNIST image classification and IMDB sentiment analysis with BERT. The assessment spans six key criteria—installation ease, configuration flexibility, interoperability, code intrusiveness, result interpretability, and documentation quality—and incorporates a weighted scoring mechanism. By establishing a balanced evaluation framework that integrates both quantitative and qualitative insights, this work delivers a clear and practical guide for selecting MLOps tools tailored to diverse application scenarios.

This study addresses the widespread neglect of licensing terms and regulatory compliance in the deployment of machine learning models within open-source software, particularly in safety-critical contexts where associated risks are pronounced. The authors present the first systematic investigation of ML usage across 173 open-source projects on GitHub spanning 16 application domains. Through code inspection and contextual analysis, they evaluate each model’s role in decision-making, the presence of risk-mitigation strategies, and adherence to licensing requirements. The findings reveal that certain projects employ ML for high-stakes decisions without complying with applicable license conditions and often lack essential post-processing safeguards. This work uncovers critical compliance blind spots in the open-source ecosystem and provides an empirical foundation for developing compliance guidelines and automated detection tools.

This study addresses the widespread absence of systematic instruction on building, testing, deploying, and maintaining AI/ML systems in current undergraduate software engineering (SE) curricula. It presents the first comprehensive delineation of core AI/ML topics essential for SE practice, integrating curriculum mapping analysis, instructor needs surveys, and structured modeling to identify critical content gaps in existing programs. Grounded in empirical evidence, the work proposes actionable pathways for embedding high-priority AI/ML themes into established SE courses. The resulting framework offers a practical, implementable guide to enhance SE education’s capacity to support the development of intelligent software systems.

This study addresses collaboration and communication (CoCo) challenges faced by machine learning engineering teams in the hardware-dominated semiconductor industry, where ambiguous roles, stringent data governance, protracted development cycles, and tight coupling with physical processes significantly impede system deployment, reproducibility, and maintenance. Focusing for the first time on this hardware-intensive domain, the research draws on semi-structured interviews with twelve practitioners from globally leading semiconductor firms, combined with thematic analysis and an interdisciplinary collaboration framework, to systematically identify sixteen recurrent CoCo challenge categories—among which role and responsibility ambiguity emerges as the most critical. The work further distills multiple empirically validated mitigation strategies, offering an evidence-based foundation and targeted recommendations for optimizing ML engineering collaboration tools and workflows in semiconductor manufacturing contexts.