pipelines

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.

Existing ETL pipelines heavily rely on manual, context-sensitive design of transformation logic, resulting in poor generalizability and low reusability. To address this, we propose an example-driven autonomous ETL framework: given user-provided target data examples, it constructs a paired-sample-based planning engine that automatically infers and synthesizes high-fidelity, context-adapted data transformation programs. Integrated with modular ETL components and runtime monitoring, the framework enables end-to-end automation for multi-format, multi-structured, and multi-scale data processing. Experiments across 14 real-world, cross-domain datasets demonstrate that our approach substantially reduces human intervention while achieving high-precision transformations (average F1 score of 0.92), strong generalization across diverse schemas and formats, and practical engineering deployability.

This work addresses the unreliability of developer productivity dashboards, which often stems from ad hoc scripts that introduce undetected silent data gaps, eroding organizational trust. To resolve this, we propose a robust ELT pipeline grounded in DAG-based orchestration and the Medallion architecture, decoupling data extraction from transformation to preserve the immutability of raw data. Our approach introduces a state-driven dependency scheduling mechanism and, for the first time, treats metric pipelines as production-grade distributed systems. We emphasize the critical role of immutable raw history in enabling reliable metric redefinition. This methodology significantly enhances data reliability and freshness while effectively eliminating silent failures, thereby restoring organizational confidence in DevOps metrics.

Ad hoc SQL development lacks engineering rigor, leading to data silos, logical redundancy, and ineffective data governance. Method: This paper proposes a DataOps-driven CI/CD framework for analytical SQL warehouses, featuring a novel five-stage automated pipeline—Lint, Optimize, Parse, Validate, Observe—that embeds quality assurance and enables end-to-end lifecycle governance. Contribution/Results: We introduce the DataOps Controls Scorecard and a requirements traceability matrix, explicitly mapping 12 governance criteria to CI/CD stages to ensure control completeness and scalability. The framework integrates Agile, Lean, and DevOps principles with static analysis, syntactic parsing, optimization recommendations, validation testing, and observability. Empirical evaluation demonstrates significant improvements in data quality, development transparency, and cross-functional collaboration, providing a sustainable, production-ready pathway for large-scale analytical systems.

This study presents the first empirical investigation into the evolution of CI/CD configurations in machine learning (ML) projects. Addressing the lack of understanding regarding how CI/CD configurations co-evolve with ML components, the authors analyze 508 open-source ML projects, 343 manually annotated commits, and 15,634 automated CI/CD commits. They propose a novel 14-category taxonomy capturing synergistic changes between CI/CD and ML components, develop a dedicated clustering tool to identify recurrent evolutionary patterns, and establish an empirically grounded model linking developer experience to CI/CD configuration modification behavior. Results show that 61.8% of CI/CD-related commits involve build strategy modifications; common anti-patterns—including dependency hardcoding and missing test frameworks—are identified; and senior developers modify CI/CD configurations more frequently and effectively than juniors, confirming the critical role of experience in CI/CD maintenance.

This work addresses the limitations of existing CI/CD workflow analyses, which often focus narrowly on stage identification and struggle to assess reliability, maintainability, and optimization priorities. To overcome this, we propose a large language model–based CI/CD analysis pipeline that integrates repository context enhancement, anti-pattern detection, stage mining, and actionable recommendation generation. Our approach uniquely combines diagnostic reasoning, context awareness, and human-in-the-loop review to deliver observability tailored to cybersecurity engineering. Leveraging few-shot prompting, YAML parsing, and statistical tests (chi-square and Cramér’s V), the method identifies 434,769 anti-patterns across 75,201 workflows and generates an average of 8.25 syntactically valid optimization suggestions per repository, achieving a 96.1% compliance rate with YAML syntax standards.

This work addresses the heavy reliance on expert knowledge in designing and debugging scientific workflows, a challenge exacerbated by existing large language model approaches that directly generate code without ensuring transparency, reproducibility, or seamless system integration. To overcome these limitations, we propose an AI-assisted scientific workflow management framework that decouples user intent from implementation through a structured specification phase, enabling specification-driven workflow generation and validation. We further introduce a multi-layer debugging agent powered by large language models to automate error diagnosis and correction. By deeply integrating with the Pegasus workflow system via the Model Context Protocol (MCP), our approach supports end-to-end workflow lifecycle management. Empirical evaluation demonstrates successful generation and execution of federated learning medical imaging workflows comprising thousands of tasks, substantially reducing debugging effort and empowering non-expert users to construct complex workflows adhering to expert-level design patterns.

This study addresses the challenges of high latency, unstable concurrency, and security risks faced by large language model (LLM) agents in automating asset lifecycle management within Industry 4.0. The authors propose a Plan-then-Execute architecture that generates verifiable workflow graphs and integrates a topology-aware parallel scheduling mechanism to enable controlled inference overlap while ensuring functional correctness and security. Key technical contributions include topological-sort-based multi-agent scheduling, structured context pruning, dependency-aware concurrency control, and graceful degradation under fault injection. Evaluated on the AssetOpsBench benchmark, the system reduces median end-to-end latency by 1.6× (up to 1.8× for highly parallel tasks) and cuts inference overhead by approximately 30% through context pruning, all while maintaining stable task completion rates and output quality.

Existing visual analytics workflows are predominantly described in unstructured textual form, hindering systematic comparison, reuse, and practical guidance. This work proposes ATWL, a formal, declarative language for modeling visual analytics workflows through a modular ontology grounded in eight artifact types and standardized intents. For the first time, this approach enables structured, machine-interpretable representations of such workflows. Leveraging large language models, the authors automatically extract workflows from academic papers to construct a reusable repository comprising 17 annotated instances. Empirical evaluation demonstrates that ATWL effectively uncovers cross-workflow structural patterns and yields more compact, structured, and extensible analytical recommendations than original narrative descriptions, thereby facilitating efficient in-context reuse and adaptation.

This work addresses the growing complexity of CI/CD pipelines and the lack of structured analysis capabilities in existing tools for understanding their behavior, failures, and version evolution. The authors propose an innovative approach that uniquely integrates digital twin technology with BPMN-based modeling in DevOps contexts. By automatically parsing raw CI configurations and execution logs, the method constructs structured, high-level process models that enable pipeline visualization, failure traceability, and cross-version comparison. Evaluated across multiple open-source projects, the approach demonstrates effectiveness in monitoring, evolutionary analysis, and fault diagnosis, offering a modular and extensible foundational framework for the analysis and optimization of CI/CD pipelines.