Key to the brand new providing is the F5 AI Gateway, which is a container or standalone package deal that runs alongside an AI software or giant language mannequin (LLM) that manages, orchestrates, and secures the interplay between numerous AI providers, knowledge sources, and user-facing purposes. It ensures site visitors administration and communication between on-premises methods, cloud-based AI fashions, edge gadgets, and APIs.
The F5 AI Gateway presents an automatic manner for purchasers to safe and handle interactions amongst AI purposes, APIs, and LLMs. It’s a containerized Kubernetes service that may be deployed by itself or built-in with current F5 software program, {hardware}, or providers, the corporate acknowledged. The gateway helps standard AI fashions resembling OpenAI, Anthropic, and Ollama in addition to generic HTTP upstream LLMs and small language mannequin (SLM) providers.
F5 has additionally added an AI assistant for its NGINX One SaaS-based software help administration console. The device is powered by the F5 AI Knowledge Material and serves as an clever accomplice to stretched NetOps, SecOps, DevOps, and platform ops groups. The AI assistant for NGINX One makes use of a pure language interface to streamline operations and helps clients configure and optimize software supply, preemptively deal with threats, and establish anomalies earlier than they impression manufacturing, based on F5.
F5 stated it should add an AI assistant for its BIG-IP software supply system. Clients will be capable to automate the creation, upkeep, and optimization of iRules whereas decreasing the time and sources required to handle site visitors and securely ship apps, F5 acknowledged.
Along with the software program, F5 has expanded its Velos {hardware} household by including a CX1610 chassis and BX520 blade. The Velos CX1610 chassis and BX520 400-Gbps blade scale to multi-terabits of throughput. The F5 Velos options knowledge packet routing, granular safety, load balancing, and low latency, enabling knowledge ingestion and real-time knowledge wants for AI workloads, F5 acknowledged.