Generative AI for business applications, such as autocompleting information and performing calculations, is a topic of interest for the Linux Foundation and industry leaders like Cloudera and Intel. The Linux Foundation recently introduced the Open Platform for Enterprise AI (OPEA) project to promote the development of open, modular generative AI systems. This initiative, overseen by the Linux Foundation’s LFAI and Data organization, aims to create robust and scalable AI systems that leverage open-source innovations.
OPEA, a Sandbox Project of the Linux Foundation, has garnered support from major industry players including IBM-owned Red Hat, Hugging Face, Domino Data Lab, MariaDB, and VMWare. The project’s focus includes enhancing support for AI toolchains, enabling AI workloads to run on various hardware components, and developing pipelines for retrieval-augmented generation (RAG).
RAG is gaining traction in business applications of relational AI due to its ability to expand a model’s information access beyond its training data. By referencing external data sources like proprietary business data or public databases, RAG models can provide more comprehensive and contextually relevant responses.
Intel, a key contributor to OPEA, highlights the project’s aim to establish standards for open and interoperable RAG solutions. OPEA seeks to streamline the selection and deployment of RAG components by collaborating with industry partners to consolidate frameworks and research solutions.
OPEA’s GitHub repository outlines a framework for evaluating relational AI systems based on performance, functionality, reliability, and enterprise readiness. The project also plans to engage with the open-source community to assess conceptual AI deployments and explore open model development opportunities.
With members like Cloudera, Domino, and VMWare investing in enterprise generative AI tooling, the potential for collaboration under OPEA is promising. By working together to create cross-compatible AI tools, vendors can offer customers a diverse range of options tailored to their specific requirements. However, the challenge lies in avoiding vendor lock-in and fostering a collaborative ecosystem for the advancement of generative AI technologies.
