Powering advanced AI data delivery with F5 and Scality

Yuichi MiyazakiDirector, Technical Alliances | F5

As organizations scale AI initiatives from experimentation to production, one reality is becoming clear: AI success depends as much on data delivery as it does on models.

Training, fine-tuning, analytics, and retrieval-augmented generation (RAG) all rely on fast, reliable access to massive volumes of data—often stored as objects and distributed across sites, environments, and regions.

S3-compatible object storage has emerged as a foundational layer for these workloads, as we discussed in a recent F5 blog series on AI data delivery. But storage alone is not enough. Without intelligent traffic management, availability controls, and security at the access layer, AI pipelines quickly encounter bottlenecks, resilience gaps, and governance challenges that slow innovation and increase risk.

This is why F5 continues to invest in an open ecosystem of storage and data-platform partners, helping customers securely deliver S3 traffic at scale, regardless of where data resides. As part of that strategy, F5 today is announcing an expanded partnership with Scality, bringing together the San Francisco-based company’s enterprise-grade, S3-compatible object storage with F5’s application delivery and security capabilities.

This partnership benefits all parties. For customers, it delivers a validated architecture for resilient, high-performance AI data pipelines across on-premises, hybrid, and multi-site environments.

For Scality, the collaboration extends the reach of its storage platform with advanced traffic management, global access policies, and integrated security services that simplify large-scale, distributed deployments. For F5, it reinforces a platform-agnostic approach to AI data delivery, ensuring consistent performance, availability, and protection across a growing range of S3-based environments.

Why S3 remains the backbone of AI workloads

S3-compatible object storage has become the lingua franca of modern AI and analytics workloads for good reason. Its flat namespace, API-driven access model, and ability to scale across petabytes of data make it well suited for the parallel, data-hungry access patterns common in AI training and analytics pipelines.

As organizations increasingly standardize on S3 semantics—whether on-premises, in the cloud, or at the edge—the ability to deliver S3 traffic efficiently and securely becomes a critical architectural requirement.

Scality’s RING object storage platform was designed with this scale in mind, providing durable, distributed storage that can span multiple sites and regions. F5 complements this with the F5 Application Delivery and Security Platform (ADSP), including the BIG-IP platform, which delivers advanced traffic management, availability, and security services at the access layer.

Together, they address a common challenge facing enterprise AI teams: how to make large-scale object storage usable, predictable, and secure for real-world AI workloads.

Why F5 is a fit for Scality

A key reason Scality partnered with F5 is F5’s long-standing expertise in application delivery and load balancing. AI and analytics workloads rarely generate smooth, predictable traffic. Instead, they produce bursts of parallel requests, uneven read and write patterns, and shifting access demands across sites and regions. Managing this traffic effectively requires more than simple round-robin distribution.

The F5 BIG-IP platform offers a broad range of load-balancing and traffic-management algorithms, along with policy-driven controls that help ensure data requests are routed efficiently and reliably. This becomes especially important in multi-site environments, where organizations may need to direct users or workloads to the optimal storage location based on proximity, performance, or policy requirements.

Scality’s global footprint, particularly its strong presence in Europe and EMEA, also makes multi-site traffic management a strategic priority. F5’s ability to support sophisticated, policy-based routing across data centers enables Scality customers to simplify access to distributed storage while maintaining consistent performance and availability.

Supported use cases for AI and data-driven environments

Together, F5 and Scality support a range of enterprise use cases that reflect how organizations are building and operating AI data pipelines today. These include:

• AI and ML training pipelines
  Training and fine-tuning models requires high-throughput, low-latency access to large datasets. F5 helps ensure that S3 requests are distributed intelligently across Scality storage nodes, reducing hot spots and improving overall pipeline efficiency.
• Multi-site data protection and disaster recovery
  Many organizations replicate object data across sites for resilience and recovery. F5’s traffic management capabilities help abstract the underlying complexity, providing a stable access layer while Scality manages replication and durability.
• Hybrid and multicloud storage architectures
  As data spans on-premises and cloud environments, consistent access becomes harder to manage. F5 provides a unified delivery and security layer for S3 traffic, helping organizations operate hybrid architectures without rewriting applications.
• Long-term data retention feeding future AI models
  AI data often has long-term value. Historical datasets used for analytics or future model training must remain accessible, protected, and performant over time. The combined solution supports secure, scalable access to these repositories as AI initiatives evolve.

Making security operational, not bolted on

AI data is both high-value and long-lived, making it an attractive target for attackers. Risks such as “harvest now, decrypt later” attacks underscore the importance of protecting sensitive datasets today—even if the threat may not materialize for years.

By centralizing TLS, access controls, and DDoS protection at the delivery layer, F5 reduces operational complexity while helping organizations enforce consistent security policies across environments. Rather than treating security as an afterthought, the joint architecture makes it an integral part of how data is delivered to AI workloads.

Why this partnership matters now

AI adoption is accelerating faster than many infrastructure assumptions were designed to support. At the same time, S3 has become the common language for storing and accessing the data that fuels AI. Organizations need validated architectures that scale, rather than custom integrations that add risk and operational overhead.

The expanded partnership between F5 and Scality offers enterprises a practical path forward—combining scalable object storage with proven application delivery and security services to support AI workloads today and into the future.

To learn more about this collaboration, read the press release and check out the technical article on the F5 DevCentral Community site.

About the Author

Yuichi MiyazakiDirector, Technical Alliances | F5

More blogs by Yuichi Miyazaki