Addressing RAG Challenges with the F5 and Red Hat Ecosystem
Generative AI’s promise of increased productivity and business empowerment has captured the attention of business leaders across many industries: According to the F5 2025 State of Application Strategy Report, virtually every respondent (99%) reported feeling comfortable using AI not merely to support decision making but to automate at least one operational function.
“Organizations everywhere are on an accelerated journey to deploy generative AI and other emerging technologies to transform how they work, innovate, and engage with their customers,” said Kunal Anand, Chief Innovation Officer at F5.
However, standard generative AI models based on off-the-shelf large language models (LLMs) often lack the up-to-date information and contextual intelligence to play a strategic, trustworthy role in today’s dynamic business environment, where real-time data access, domain-specific expertise, and accurate decision-making are critical for success.
Organizations seeking to integrate generative AI applications with a source for more up-to-date, accurate, and domain-specific information are deploying retrieval-augmented generation (RAG), an AI framework that combines retrieval-based and generative approaches to improve the quality and accuracy of AI-generated responses.
Enterprise AI is a team sport
Successfully implementing RAG into an enterprise AI infrastructure isn’t simple plug-and-play. RAG requires orchestration of multiple technology components and technologies, and is an excellent example of how enterprise AI solutions require an open and collaborative ecosystem of technology providers, infrastructure partners, and channel partners to succeed.
Deploying RAG involves multiple vendors, including providers of data storage, container platforms, LLMs, traffic management tools, APIs and backend infrastructure, and more. Enterprise AI is truly a team sport, and at F5 we recognize that many players must come together to make AI at scale possible for our customers. Success requires the power of a collaborative network that includes tech providers to deliver best-in-class components and software, channel partners to translate tools into business outcomes, and a mindset that sees AI not as a product, but as a solution ecosystem.
A superior foundation for AI
F5 and Red Hat OpenShift AI provide a robust, unified foundation for deploying RAG into LLMs, addressing the significant security and traffic management complexities involved in integrating data from multiple sources into a generative AI system. Red Hat and F5 have a long history of technology partnership, and the companies’ current collaboration delivers a robust, protected, and scalable platform for AI-driven initiatives, including RAG integration into existing LLMs.
F5 and Red Hat are proven leaders: F5 has been named one of the 25 “hottest AI companies for data center and edge” by CRN, citing F5’s “all-in-one application delivery and security platform … to help enterprises address multi-cloud networking, AI and API security demands.” Red Hat is recognized as a leader in open source enterprise solutions by Dataquest, stating that its “open-source innovation will define the next era of enterprise IT.”
The combination of F5’s advanced security and traffic management technologies and Red Hat OpenShift AI, an advanced AI application development platform, help establish a more secure and scalable structure for AI-driven initiatives. Organizations can confidently deploy their AI applications, knowing they are protected against evolving cyber threats and capable of meeting the demanding requirements of modern AI workloads.
Red Hat and F5 and its ecosystem of technology partners will continue to work together to help customers connect and secure their apps and APIs in complex, hybrid environments. As organizations explore the potential of AI, they can rely on trusted vendors like Red Hat and F5 to deliver a consistent, unified platform to keep AI environments manageable, cost-efficient, and secure, fostering an environment where innovation and security go hand in hand.
Challenges of integrating RAG into existing LLMs
RAG is increasingly essential for enterprise-grade generative AI applications, and the RAG market is poised for substantial growth in the coming years, with a compound annual growth rate (CAGR) of 44.7% from 2024 to 2030, potentially reaching $11.03 billion by 2030.
While RAG can greatly enhance the outputs of generative AI applications, deploying RAG into LLMs is not straightforward, as it greatly increases system complexity and security risks due to the fact that RAG’s additional data components exist in multiple digital environments and external knowledge bases. The necessary network connections required to pass retrieval requests and the augmented data responses bring security and data-leakage hazards and require stringent access controls and encryption mechanisms to protect sensitive data.
Integrating Red Hat OpenShift AI with F5 Distributed Cloud Services and F5 BIG-IP Next for Kubernetes provides a bedrock set of services that deliver advanced security and traffic management technologies to support and protect multiple RAG use cases.
Distributed Cloud Services are SaaS-based security, networking, and application management services that enable customers to deploy, secure, and operate their applications in a cloud-native environment wherever needed—in the data center, in multi-cloud environments, or the network or enterprise edge.
In turn, BIG-IP Next for Kubernetes optimizes traffic management and distribution across AI-processing clusters, ensuring scalable and efficient API communication to support RAG-based workloads, ensuring reliable and high-performance data flows without performance degradation.
Meta’s recent announcement regarding its Llama 4 herd of LLMs, with Llama 4 Scout featuring a 10 million token context window, has stirred commentary that expansive context windows like these could soon render RAG obsolete. However, Llama 4 LLMs are pretrained and self-contained, without real-time access to external or private data unless explicitly designed for those features. Conversely, RAG enables models fetch and incorporate the most current and relevant information from vast data stores in real time, ensuring AI outputs are both accurate and contextually relevant. F5 believes RAG will continue to be a foundational generative AI mechanism because it better accommodates the dynamic and interactive nature of AI workflows.
A joint F5 and Red Hat platform for protecting RAG workloads
Deploying Distributed Cloud Services and BIG-IP Next for Kubernetes on Red Hat OpenShift AI creates a fortified security solution for protecting RAG workloads. The joint F5 and Red Hat platform enhances RAG reliability and scalability, providing an easily managed foundation for RAG-empowered generative AI solutions that deliver increased accuracy and relevance for improved outcomes and a better user experience.
For more information, read the solution overview. And if you’re planning to be at this year’s Red Hat Summit, please attend an F5 session.