In today’s data-driven era, the vision of the CyclOps project is to facilitate the adoption and production of data, models, and services from and for data spaces to enable AI-based data-driven applications for all players, business and research alike. The main objective of CyclOps is the operationalisation and automation of the entire data life cycle, ensuring that data remains FAIR—Findable, Accessible, Interoperable, and Reusable. CyclOps’ novelty resides in its Integrated Knowledge Base, which leverages Knowledge Graphs (KGs) as formal models to represent data and metadata while providing context and guaranteeing interoperability.
CyclOps’ spans technical and societal challenges. On the one hand, it aims to address the entire data life cycle from data generation to processing and final disposal while guaranteeing interoperability and portability of data between and across data spaces. On the other hand, CyclOps must consider cybersecurity, privacy, and fairness aspects at its core, as well as provide means to enforce data-related rights, obligations, and responsibilities while addressing sociocultural factors.
The first step is operationalising the data life cycle to achieve the envisioned platform. This is the definition of a cohesive platform and the definition of the data processes and flows that implement its principles. For that, CyclOps will consider pipelines composed of DataOps and AIOps operations. The DataOps layer will enable data management tasks like discovery, curation, quality, and integration. At the same time, AIOps provides a decentralised repository of AI algorithms and makes the resulting AI models available. The execution of such pipelines is carried out by the Data and Execution Abstraction layer, which facilitates the development of distributed data preprocessing and analysis. All these components will benefit from the Integrated Knowledge Base within the Data Governance and Trust layer, which will provide mechanisms for traceability, explainability, cybersecurity and privacy.
The automation of the data life cycle involves allowing developers to avoid low-level technical details when implementing pipelines. Thus, CyclOps will automatically deploy pipelines that satisfy the user requirements bridging the gap between non-technical users and automated systems via the Intent-based Human Interface. This user-centric design allows for intuitive interaction with data spaces, making data exploration and visualization accessible to non-technical users.
The CyclOps project will be validated on 4 use cases on the topics of tourism, green deal, procurement and manufacturing, spanning 5 European data spaces. Overall, CyclOps will not only enhancee data integration and analysis but also ensures that data remains secure, compliant, and accessible bringing efficiency and innovation to SMEs.
