pith. sign in

arxiv: 2512.16455 · v3 · submitted 2025-12-18 · 💻 cs.DC · cs.AI

AI4EOSC: a Federated Cloud Platform for Artificial Intelligence in Scientific Research

Ignacio Heredia , \'Alvaro L\'opez Garc\'ia , Fernando Aguilar G\'omez , Diego Aguirre , Caterina Alarc\'on Mar\'in , Khadijeh Alibabaei , Lisana Berberi , Miguel Caballer
show 23 more authors
Amanda Calatrava Pedro Castro Alessandro Costantini Mario David Jaime D\'iez Stefan Dlugolinsky Borja Esteban Sanchis Giacinto Donvito Leonhard Duda Sa\'ul Fernandez Andr\'es Heredia Canales Valentin Kozlov Sergio Langarita Jo\~ao Machado Germ\'an Molt\'o Daniel San Mart\'in Martin \v{S}eleng Giang Nguyen Marcin P{\l}\'ociennik Marta Obreg\'on Ruiz Susana Rebolledo Ruiz Vicente Rodriguez Judith S\'ainz-Pardo D\'iaz Viet Tran
This is my paper

Pith reviewed 2026-05-16 21:29 UTC · model grok-4.3

classification 💻 cs.DC cs.AI
keywords federated cloud platformAI/ML lifecycleEOSCFAIR principlesMLOpsopen scienceserverless computingprovenance tracking
0
0 comments X

The pith

AI4EOSC is a federated open-source platform that runs the full AI/ML lifecycle inside the European Open Science Cloud using modular architecture and built-in FAIR metadata.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents AI4EOSC as a way to close the gap between ordinary MLOps tools and the requirements of open scientific research. It builds a federated platform that combines an AI development environment, a serverless service layer, and orchestration across separate research infrastructures. The system adds FAIR-by-design metadata using MLDCAT-AP and W3C PROV provenance inside its CI/CD pipeline. This setup matters because it cuts the manual work researchers spend on infrastructure and makes models easier to find, reuse, and reproduce across sites. The authors show the approach works through multiple community installations and real scientific cases.

Core claim

AI4EOSC is a federated, open-source platform that operationalizes the full AI/ML lifecycle within the EOSC ecosystem. It uses a modular and distributed architecture that includes an AI development platform, a serverless AI-as-a-Service layer, and a federated orchestration model to integrate heterogeneous compute and storage resources. The platform enforces FAIR principles by standardizing metadata with MLDCAT-AP and tracking provenance according to W3C PROV through an integrated CI/CD pipeline. Its value is demonstrated by consistent deployments across heterogeneous cloud providers and validation through scientific cases that show reduced manual burden and improved reproducibility.

What carries the argument

Modular distributed architecture with federated orchestration that pulls together heterogeneous resources while enforcing MLDCAT-AP metadata and W3C PROV provenance tracking.

Load-bearing premise

The assumption that the modular architecture can integrate compute and storage resources from different providers in a consistent and seamless way without major compatibility or performance problems.

What would settle it

A documented case where an AI workflow fails to deploy or shows large performance differences when run on resources from two or more distinct cloud providers in the same installation.

Figures

Figures reproduced from arXiv: 2512.16455 by Alessandro Costantini, \'Alvaro L\'opez Garc\'ia, Amanda Calatrava, Andr\'es Heredia Canales, Borja Esteban Sanchis, Caterina Alarc\'on Mar\'in, Daniel San Mart\'in, Diego Aguirre, Fernando Aguilar G\'omez, Germ\'an Molt\'o, Giacinto Donvito, Giang Nguyen, Ignacio Heredia, Jaime D\'iez Stefan Dlugolinsky, Jo\~ao Machado, Judith S\'ainz-Pardo D\'iaz, Khadijeh Alibabaei, Leonhard Duda, Lisana Berberi, Marcin P{\l}\'ociennik, Mario David, Marta Obreg\'on Ruiz, Martin \v{S}eleng, Miguel Caballer, Pedro Castro, Sa\'ul Fernandez, Sergio Langarita, Susana Rebolledo Ruiz, Valentin Kozlov, Vicente Rodriguez, Viet Tran.

Figure 1
Figure 1. Figure 1: Overview of the AI4EOSC architecture the federation and contribute to the existing AI4EOSC platform. 3.3. Workload Management System The Workload Management System (WMS) is a fed￾eration of service providers that is used to run the platform jobs. Deployed with Ansible [43], it uses Consul [44] to federate the providers (following a WAN approach) and Nomad [45] to run the work￾loads. The current federation … view at source ↗
Figure 2
Figure 2. Figure 2: CI/CD pipeline for AI modules 3.7. Authentication At the PaaS layer the AI4EOSC Platform adopts the INDIGO Identity and Access Management (INDIGO￾IAM) [84] service, in line with the use of other INDIGO components at the PaaS layer (§3.2). It features attribute and identity management (to manage group member￾ship, attributes assignment and account linking), as well as Multi Factor Authentication integration… view at source ↗
read the original abstract

The rapid growth of Artificial Intelligence and Machine Learning in scientific research has highlighted a gap between industry-standard MLOps tools and platforms, and the unique requirements of modern and Open Science, particularly regarding the FAIR (Findable, Accessible, Interoperable, and Reusable) principles. This paper presents AI4EOSC, a federated, open-source platform designed to operationalize the full AI/ML lifecycle within the European Open Science Cloud (EOSC) ecosystem. Our methodology tackles the fragmentation of distributed research infrastructures by integrating a modular and distributed architecture comprising an AI development platform, a serverless AI-as-a-Service layer, and a federated orchestration model that is able to integrate heterogeneous compute and storage resources from distributed e-Infrastructures. AI4EOSC also introduces a ``FAIR-by-design'' approach that enforces metadata standardization (via MLDCAT-AP) and W3C PROV-compliant provenance tracking through a platform-integrated CI/CD pipeline. AI4EOSC added value is demonstrated through the delivery of a diverse set of community installations, showing consistent and seamless deployment across heterogeneous cloud providers. These installations are validated by a set of scientific cases, showing how our work reduces the manual burden on researchers while ensuring high levels of reproducibility and interoperability and providing an unified environment for development, training, and production of AI/ML models in the EOSC.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The manuscript presents AI4EOSC, a federated open-source platform for operationalizing the full AI/ML lifecycle within the European Open Science Cloud (EOSC). It integrates a modular distributed architecture (AI development platform, serverless AI-as-a-Service layer, federated orchestration for heterogeneous compute/storage resources), a FAIR-by-design approach enforcing MLDCAT-AP metadata and W3C PROV provenance via CI/CD pipelines, and demonstrates value through community installations and scientific use cases that reduce researcher burden while improving reproducibility and interoperability.

Significance. If the seamless integration of heterogeneous resources holds, the work would address fragmentation in distributed research infrastructures and provide a standards-compliant environment for AI in open science, advancing reproducibility and FAIR compliance within EOSC. The open-source nature and emphasis on provenance tracking represent practical strengths for community adoption.

major comments (1)
  1. [Abstract and validation description] The central claim of 'consistent and seamless deployment across heterogeneous cloud providers' (Abstract) rests on descriptive accounts of installations and cases without quantitative metrics such as deployment success rates, performance overhead, compatibility failure modes, or cross-provider variance. This leaves the assumption of robust federated orchestration unverified by measurement.
minor comments (1)
  1. [Abstract] The abstract is information-dense; separating the architecture description from the validation claims would improve readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the positive evaluation of AI4EOSC's significance and for the constructive feedback on validation. We address the major comment below and describe the revisions planned for the manuscript.

read point-by-point responses

  1. Referee: [Abstract and validation description] The central claim of 'consistent and seamless deployment across heterogeneous cloud providers' (Abstract) rests on descriptive accounts of installations and cases without quantitative metrics such as deployment success rates, performance overhead, compatibility failure modes, or cross-provider variance. This leaves the assumption of robust federated orchestration unverified by measurement.

    Authors: We acknowledge that the current manuscript validates the federated orchestration primarily through descriptive accounts of successful community installations across heterogeneous providers and their use in scientific cases. These demonstrate operational functionality and seamlessness in practice, but we agree that the absence of explicit quantitative metrics (e.g., success rates, overhead, or variance) leaves the robustness claim less strongly evidenced than it could be. In the revised manuscript we will add a dedicated subsection under validation reporting available quantitative indicators from our deployment logs and CI/CD pipelines, including deployment success rates, average provisioning times, and any observed compatibility notes across providers. This will provide measurable support while preserving the paper's primary focus on architecture, serverless integration, and FAIR-by-design provenance. revision: yes

Circularity Check

0 steps flagged

No circularity: platform description and empirical validation are self-contained

full rationale

The paper presents a descriptive account of a federated platform architecture, its modular components, serverless layer, orchestration model, and FAIR-by-design metadata approach. Validation rests on reported community installations and scientific use cases rather than any mathematical derivation, fitted parameters, or predictions. No equations, self-definitional constructs, or load-bearing self-citations appear in the derivation chain; the central claims concern design choices and deployment feasibility, which do not reduce to their own inputs by construction. This is the expected non-finding for an engineering/platform paper without quantitative modeling.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is an infrastructure and platform paper with no mathematical derivations, fitted parameters, or postulated physical entities; the central claims rest on engineering integration of existing technologies.

pith-pipeline@v0.9.0 · 5715 in / 1050 out tokens · 53666 ms · 2026-05-16T21:29:11.442813+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

131 extracted references · 131 canonical work pages · 3 internal anchors

  1. [1]

    doi:10.48550/arXiv.2502.03544 , url =

    Y . Chervonyi, T. H. Trinh, M. Olšák, X. Yang, H. Nguyen, M. Menegali, J. Jung, V . Verma, Q. V . Le, T. Luong, Gold- medalist performance in solving olympiad geometry with al- phageometry2, arXiv preprint arXiv:2502.03544 (2025)

    work page arXiv 2025

  2. [2]

    Jumper, R

    J. Jumper, R. Evans, A. Pritzel, T. Green, M. Figurnov, O. Ronneberger, K. Tunyasuvunakool, R. Bates, A. Žídek, A. Potapenko, et al., Highly accurate protein structure pre- diction with alphafold, nature 596 (2021) 583–589

    work page 2021

  3. [3]

    Espeholt, S

    L. Espeholt, S. Agrawal, C. Sønderby, M. Kumar, J. Heek, C. Bromberg, C. Gazen, R. Carver, M. Andrychowicz, J. Hickey, et al., Deep learning for twelve hour precipitation forecasts, Nature communications 13 (2022) 5145

    work page 2022

  4. [4]

    H. O. Khogali, S. Mekid, The blended future of automation and ai: Examining some long-term societal and ethical impact features, Technology in Society 73 (2023) 102232. URL:https://www.sciencedirect.com/ science/article/pii/S0160791X23000374. doi:https: //doi.org/10.1016/j.techsoc.2023.102232

    work page doi:10.1016/j.techsoc.2023.102232 2023

  5. [5]

    Sergeyuk, Y

    A. Sergeyuk, Y . Golubev, T. Bryksin, I. Ahmed, Using ai- based coding assistants in practice: State of affairs, percep- tions, and ways forward, Information and Software Tech- nology 178 (2025) 107610. URL:http://dx.doi.org/10. 1016/j.infsof.2024.107610. doi:10.1016/j.infsof. 2024.107610

    work page doi:10.1016/j.infsof 2025

  6. [6]

    Q. Hao, F. Xu, Y . Li, J. Evans, Ai expands scientists’ impact but contracts science’s focus, 2024. URL:https://arxiv. org/abs/2412.07727.arXiv:2412.07727

    work page arXiv 2024

  7. [7]

    Calatrava, H

    A. Calatrava, H. Asorey, J. Astalos, A. Azevedo, F. Ben- incasa, I. Blanquer, M. Bobak, F. Brasileiro, L. Codó, L. del Cano, B. Esteban, M. Ferret, J. Handl, T. Kerzen- macher, V . Kozlov, A. K ˇrenek, R. Martins, M. Pavesio, A. J. Rubio-Montero, J. Sánchez-Ferrero, A survey of the european open science cloud services for ex- panding the capacity and capa...

    work page doi:10.1016/j.cosrev.2023.100571 2023

  8. [8]

    EOSC Association, European Commission, The eosc part- nership monitoring framework v7.0,https://eosc.eu/ wp-content/uploads/2024/08/20240621_Draft_ EOSC_Monitoring_Framework_v7_Final_Clean.pdf,

    work page 2024

  9. [9]

    Version 7.0, Final Clean Draft

    work page

  10. [10]

    M. D. Wilkinson, M. Dumontier, I. J. Aalbersberg, G. Apple- ton, M. Axton, A. Baak, N. Blomberg, J.-W. Boiten, L. B. da Silva Santos, P. E. Bourne, et al., The fair guiding principles for scientific data management and stewardship, Scientific data 3 (2016) 1–9

    work page 2016

  11. [11]

    Ouyang, F

    W. Ouyang, F. Beuttenmueller, E. Gómez-de Mariscal, C. Pape, T. Burke, C. Garcia-López-de Haro, C. Russell, L. Moya-Sans, C. De-La-Torre-Gutiérrez, D. Schmidt, et al., Bioimage model zoo: a community-driven resource for ac- cessible deep learning in bioimage analysis, BioRxiv (2022) 2022–06

    work page 2022

  12. [12]

    Kreuzhuber, J

    Žiga Avsec, R. Kreuzhuber, J. Israeli, N. Xu, J. Cheng, A. Shrikumar, A. Banerjee, D. S. Kim, L. Urban, A. Kun- daje, O. Stegle, J. Gagneur, Kipoi: accelerating the commu- nity exchange and reuse of predictive models for genomics, bioRxiv (2018). URL:https://api.semanticscholar. org/CorpusID:92535930

    work page 2018

  13. [13]

    T. Wolf, L. Debut, V . Sanh, J. Chaumond, C. Delangue, A. Moi, P. Cistac, C. Ma, Y . Jernite, J. Plu, C. Xu, T. Le Scao, S. Gugger, M. Drame, Q. Lhoest, A. M. Rush, Transform- ers: State-of-the-Art Natural Language Processing, in: ACL Anthology, Association for Computational Linguistics, 2020, pp. 38–45. URL:https://www.aclweb.org/anthology/ 2020.emnlp-demos.6

    work page 2020

  14. [14]

    [Accessed: 2025-08-21]

    Replicate,https://replicate.com/, 2025. [Accessed: 2025-08-21]

    work page 2025

  15. [15]

    Berberi, V

    L. Berberi, V . Kozlov, G. Nguyen, J. Sáinz-Pardo Díaz, A. Calatrava, G. Moltó, V . Tran, Á. López García, Ma- chine learning operations landscape: platforms and tools, Artificial Intelligence Review 58 (2025) 167. doi:10.1007/ s10462-025-11164-3

    work page 2025

  16. [16]

    Zaharia, A

    M. Zaharia, A. Chen, A. Davidson, A. Ghodsi, S. A. Hong, A. Konwinski, S. Murching, T. Nykodym, P. Ogilvie, M. Parkhe, et al., Accelerating the machine learning lifecycle with mlflow., IEEE Data Eng. Bull. 41 (2018) 39–45

    work page 2018

  17. [17]

    Jayasiri, N

    V . Jayasiri, N. Wijerathne, A. Narasinghe, L. Nishshanke, labml.ai: A library to organize machine learning experiments,

    work page

  18. [18]

    URL:https://labml.ai/

    work page

  19. [19]

    URL:https://www

    CoreWeave, Weights and biases, 2020. URL:https://www. wandb.com/

    work page 2020

  20. [20]

    [Accessed: 2025- 08-21]

    Valohai,https://valohai.com/, 2025. [Accessed: 2025- 08-21]

    work page 2025

  21. [21]

    [Accessed: 2025- 08-21]

    Comet,https://www.comet.com/, 2025. [Accessed: 2025- 08-21]

    work page 2025

  22. [22]

    Tkachenko, M

    M. Tkachenko, M. Malyuk, A. Holmanyuk, N. Liubimov, La- bel Studio: Data labeling software, 2020-2025. URL:https: //github.com/HumanSignal/label-studio

    work page 2020

  23. [23]

    Nakayama, T

    H. Nakayama, T. Kubo, J. Kamura, Y . Taniguchi, X. Liang, doccano: Text Annotation Tool for Human, 2018. URL: https://github.com/doccano/doccano

    work page 2018

  24. [24]

    [Accessed: 2025-08-21]

    Kern AI, Refinery,https://github.com/code-kern-ai/ refinery, 2025. [Accessed: 2025-08-21]

    work page 2025

  25. [25]

    [Accessed: 2025-08- 21]

    Prodigy,https://prodi.gy/, 2025. [Accessed: 2025-08- 21]

    work page 2025

  26. [26]

    [Accessed: 2025-08-21]

    Roboflow,https://roboflow.com/, 2025. [Accessed: 2025-08-21]

    work page 2025

  27. [27]

    Fxhenn: Fpga-based acceleration framework for homomorphic encrypted cnn inference,

    CV AT.ai Corporation, Computer Vision Annotation Tool (CV AT), 2024. URL:https://doi.org/10.5281/zenodo. 12771595. doi:10.5281/zenodo.12771595

    work page doi:10.5281/zenodo 2024

  28. [28]

    [Accessed: 2025-08-21]

    H2O.ai,https://h2o.ai/, 2025. [Accessed: 2025-08-21]

    work page 2025

  29. [29]

    [Accessed: 2025-08-21]

    Alphabet, Google Vertex AI,https://cloud.google.com/ vertex-ai, 2025. [Accessed: 2025-08-21]. 12

    work page 2025

  30. [30]

    [Accessed: 2025-08-21]

    Amazon, Amazon Sagemaker,https://aws.amazon.com/ sagemaker/, 2025. [Accessed: 2025-08-21]

    work page 2025

  31. [31]

    azure.com/, 2025

    Microsoft, Azure AI Machine Learning Studio,https://ml. azure.com/, 2025. [Accessed: 2025-08-21]

    work page 2025

  32. [32]

    [Ac- cessed: 2025-08-21]

    Polyaxon, Polyaxon,https://polyaxon.com/, 2025. [Ac- cessed: 2025-08-21]

    work page 2025

  33. [33]

    Á. L. García, J. M. De Lucas, M. Antonacci, W. Zu Castell, M. David, M. Hardt, L. L. Iglesias, G. Moltó, M. Plociennik, V . Tran, et al., A cloud-based framework for machine learn- ing workloads and applications, IEEE access 8 (2020) 18681– 18692

    work page 2020

  34. [34]

    A. e. a. Costantini, A cloud-edge orchestration platform for the innovative industrial scenarios of the iotwins project, Gervasi, O., et al. Computational Science and Its Appli- cations – ICCSA 2021. ICCSA 2021. Lecture Notes in Computer Science 12950 (2021). doi:https://doi.org/10. 1007/978-3-030-86960-1_37

    work page 2021

  35. [35]

    Salomoni, I

    D. Salomoni, I. Campos, L. Gaido, J. M. de Lucas, P. Solagna, J. Gomes, L. Matyska, P. Fuhrman, M. Hardt, G. Donvito, et al., Indigo-datacloud: A platform to facilitate seamless ac- cess to e-infrastructures, Journal of Grid Computing 16 (2018) 381–408

    work page 2018

  36. [36]

    Caballer, I

    M. Caballer, I. Blanquer, G. Moltó, C. de Alfonso, Dy- namic management of virtual infrastructures, Jour- nal of Grid Computing 13 (2015) 53–70. URL:http: //link.springer.com/10.1007/s10723-014-9296-5. doi:10.1007/s10723-014-9296-5

    work page doi:10.1007/s10723-014-9296-5 2015

  37. [37]

    Caballer, G

    M. Caballer, G. Moltó, A. Calatrava, I. Blanquer, Infrastructure manager: A tosca-based orchestrator for the computing contin- uum, Journal of Grid Computing 21 (2023) 51. URL:https: //link.springer.com/10.1007/s10723-023-09686-7. doi:10.1007/s10723-023-09686-7

    work page doi:10.1007/s10723-023-09686-7 2023

  38. [38]

    Giommi, G

    L. Giommi, G. Savarese, G. Vino, D. Ranieri, A. Costantini, G. Donvito, Improving the cloud provider ranking in the indigo paas orchestration system using ai techniques, Ger- vasi, O., et al. Computational Science and Its Applications – ICCSA 2025 Workshops. ICCSA 2025. Lecture Notes in Computer Science 15886 (2025). doi:https://doi.org/10. 1007/978-3-031...

    work page 2025

  39. [39]

    Proceedings of Science 476, 1002 (2025) https://doi.org/10.22323/1

    G. Savarese, M. Antonacci, L. Giommi, Federation-registry: the renovated configuration management database for dynamic cloud federation, PoS ISGC2024 (2024). doi:10.22323/1. 458.0021

    work page doi:10.22323/1 2024

  40. [40]

    Ramírez, FastAPI, 2025

    S. Ramírez, FastAPI, 2025. URL:https://github.com/ fastapi/fastapi

    work page 2025

  41. [41]

    [Ac- cessed 12-08-2025]

    Neo4j,https://github.com/neo4j/neo4j, 2025. [Ac- cessed 12-08-2025]

    work page 2025

  42. [42]

    [Accessed 12-08-2025]

    Pallets Team, Flask,https://github.com/pallets/ flask, 2025. [Accessed 12-08-2025]

    work page 2025

  43. [43]

    [Accessed 12- 08-2025]

    MySQL,https://www.mysql.com/, 2025. [Accessed 12- 08-2025]

    work page 2025

  44. [44]

    Palma, M

    D. Palma, M. Rutkowski, T. Spatzier, TOSCA Simple Profile in Y AML Version 1.0, OASIS Committee Spec- ification Draft 04/Public Review Draft 01, OASIS, 2015-08-27. URL:http://docs.oasis-open.org/ tosca/TOSCA-Simple-Profile-YAML/v1.0/csprd01/ TOSCA-Simple-Profile-YAML-v1.0-csprd01.html

    work page 2015

  45. [45]

    [Accessed 12-08-2025]

    Red Hat, Ansible,https://docs.ansible.com/, 2025. [Accessed 12-08-2025]

    work page 2025

  46. [46]

    com/consul?product_intent=consul, 2025

    HashiCorp, Consul,https://developer.hashicorp. com/consul?product_intent=consul, 2025. [Accessed 12-08-2025]

    work page 2025

  47. [47]

    com/consul?product_intent=nomad, 2025

    HashiCorp, Nomad,https://developer.hashicorp. com/consul?product_intent=nomad, 2025. [Accessed 12-08-2025]

    work page 2025

  48. [48]

    A. Abid, A. Abdalla, A. Abid, D. Khan, A. Alfozan, J. Zou, Gradio: Hassle-free sharing and testing of ml models in the wild, arXiv preprint arXiv:1906.02569 (2019)

    work page internal anchor Pith review Pith/arXiv arXiv 1906

  49. [49]

    Traefik Labs, Traefik,https://traefik.io/traefik/,

    work page

  50. [51]

    Boyer, Docuum,https://github.com/stepchowfun/ docuum, 2025

    S. Boyer, Docuum,https://github.com/stepchowfun/ docuum, 2025. [Accessed 12-08-2025]

    work page 2025

  51. [52]

    Pérez, S

    A. Pérez, S. Risco, D. M. Naranjo, M. Caballer, G. Moltó, On- premises serverless computing for event-driven data process- ing applications, in: 2019 IEEE 12th International conference on cloud computing (CLOUD), IEEE, 2019, pp. 414–421

    work page 2019

  52. [53]

    GRyCAP Team, OSCAR Documentation: Introduction and User Guide,https://docs.oscar.grycap.net/latest/,

    work page

  53. [54]

    [Accessed: 2025-08-21]

    work page 2025

  54. [55]

    [Accessed: 2025-08- 21]

    MinIO,https://www.min.io/, 2025. [Accessed: 2025-08- 21]

    work page 2025

  55. [56]

    [Accessed 12-08-2025]

    OpenJS Foundation & Contributors, Node-RED,https:// nodered.org/, 2025. [Accessed 12-08-2025]

    work page 2025

  56. [57]

    [Accessed 12-08-2025]

    FlowFuse Inc, FlowFuse,https://flowfuse.com/, 2025. [Accessed 12-08-2025]

    work page 2025

  57. [58]

    Elyra AI, Elyra,https://github.com/elyra-ai/elyra,

    work page

  58. [60]

    com/kubernetes/kube-state-metrics, 2025

    Kubernetes Team, Kube State Metrics,https://github. com/kubernetes/kube-state-metrics, 2025. [Accessed 12-08-2025]

    work page 2025

  59. [61]

    [Accessed 12- 08-2025]

    GoAccess,https://goaccess.io/, 2025. [Accessed 12- 08-2025]

    work page 2025

  60. [62]

    [Accessed 12-08-2025]

    Prometheus,https://prometheus.io/, 2025. [Accessed 12-08-2025]

    work page 2025

  61. [63]

    [Accessed 12-08-2025]

    Cookiecutter,https://www.cookiecutter.io/, 2025. [Accessed 12-08-2025]

    work page 2025

  62. [64]

    Á. L. García, Deepaas api: A rest api for machine learning and deep learning models, Journal of Open Source Software 4 (2019) 1517

    work page 2019

  63. [65]

    [Ac- cessed 12-08-2025]

    JSON Schema,https://json-schema.org/, 2025. [Ac- cessed 12-08-2025]. 13

    work page 2025

  64. [66]

    [Accessed 12-08- 2025]

    SEMIC EU, MLDCAT-AP,https://semiceu.github.io/ MLDCAT-AP/releases/2.0.0/, 2025. [Accessed 12-08- 2025]

    work page 2025

  65. [67]

    [Accessed 12- 08-2025]

    Jenkins,https://www.jenkins.io/, 2025. [Accessed 12- 08-2025]

    work page 2025

  66. [68]

    [Accessed 12-08-2025]

    tox,https://github.com/tox-dev/tox, 2025. [Accessed 12-08-2025]

    work page 2025

  67. [69]

    van Rossum, B

    G. van Rossum, B. Warsaw, N. Coghlan, Style Guide for Python Code, PEP 8, Python Software Foundation, 2001. URL:https://www.python.org/dev/peps/pep-0008/

    work page 2001

  68. [70]

    Krekel, B

    H. Krekel, B. Oliveira, R. Pfannschmidt, F. Bruynooghe, B. Laugher, F. Bruhin, pytest,https://github.com/ pytest-dev/pytest, 2004. Contributors include Hol- ger Krekel, Bruno Oliveira, Ronny Pfannschmidt, Floris Bruynooghe, Brianna Laugher, Florian Bruhin, and others

    work page 2004

  69. [71]

    [Ac- cessed 12-08-2025]

    Bandit,https://github.com/PyCQA/bandit, 2025. [Ac- cessed 12-08-2025]

    work page 2025

  70. [72]

    URL:https://www.zenodo.org/

    European Organization For Nuclear Research, OpenAIRE, Zenodo, 2013. URL:https://www.zenodo.org/. doi:10. 25495/7GXK-RD71

    work page 2013

  71. [73]

    Abadi, P

    M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, et al., Tensor- Flow: a system for Large-Scale machine learning, in: 12th USENIX symposium on operating systems design and imple- mentation (OSDI 16), 2016, pp. 265–283

    work page 2016

  72. [74]

    PyTorch: An Imperative Style, High-Performance Deep Learning Library

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Köpf, E. Z. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, S. Chin- tala, Pytorch: An imperative style, high-performance deep learning library, CoRR abs/1912.01703 (2019). URL:http: //arxiv.o...

    work page internal anchor Pith review Pith/arXiv arXiv 1912

  73. [75]

    Microsoft, VSCode,https://code.visualstudio.com/,

    work page

  74. [77]

    [Accessed 12-08-2025]

    Coder, Code Server,https://github.com/coder/ code-server, 2025. [Accessed 12-08-2025]

    work page 2025

  75. [78]

    [Ac- cessed 12-08-2025]

    Jupyter, JupyterLab,https://jupyter.org/, 2025. [Ac- cessed 12-08-2025]

    work page 2025

  76. [79]

    Nguyen, J

    G. Nguyen, J. Sáinz-Pardo Díaz, A. Calatrava, L. Berberi, O. Lytvyn, V . Kozlov, V . Tran, G. Moltó, Á. López García, Landscape of machine learning evolution: privacy-preserving federated learning frameworks and tools, Artificial Intelligence Review 58 (2025) 51. doi:10.1007/s10462-024-11036-2

    work page doi:10.1007/s10462-024-11036-2 2025

  77. [80]

    D. J. Beutel, T. Topal, A. Mathur, X. Qiu, J. Fernandez- Marques, Y . Gao, L. Sani, K. H. Li, T. Parcollet, P. P. B. de Gusmão, et al., Flower: A friendly federated learning re- search framework, arXiv preprint arXiv:2007.14390 (2020)

    work page internal anchor Pith review arXiv 2007

  78. [81]

    H. R. Roth, Y . Cheng, Y . Wen, I. Yang, Z. Xu, Y .-T. Hsieh, K. Kersten, A. Harouni, C. Zhao, K. Lu, et al., Nvidia flare: Federated learning from simulation to real-world, arXiv preprint arXiv:2210.13291 (2022)

    work page arXiv 2022

  79. [82]

    J. S.-P. Díaz, Álvaro López García, Enhancing the conver- gence of federated learning aggregation strategies with limited data, 2025. URL:https://arxiv.org/abs/2501.15949. arXiv:2501.15949

    work page arXiv 2025

  80. [83]

    J. S.-P. Díaz, A. Athanasiou, K. Jung, C. Palamidessi, Álvaro López García, Metric privacy in federated learning for medical imaging: Improving convergence and preventing client infer- ence attacks, 2025. URL:https://arxiv.org/abs/2502. 01352.arXiv:2502.01352

    work page arXiv 2025

Showing first 80 references.