Deploy, Secure, and Optimize AI Models for Intelligent Apps with F5 and Intel

The new generation of intelligent apps uses AI automation based on data inputs—for example, an insurance app that monitors for severe weather—or it might use machine learning to adapt and improve the user experience. Gartner estimates that by 2026, 30% of new applications will use AI to drive personalized adaptive user interfaces, up from less than 5% today.² 

But many organizations lack the skills to build or use the AI models they need. According to a World Economic Forum survey, executives estimate that up to 40% of their overall workforce may need to reskill to work effectively alongside AI.³ In response, new solutions are emerging to allow developers to leverage AI models with minimal code.

AI models also require significant computing power, which can create limitations. While placing AI models close to the input source reduces latency for inference, edge locations often lack the resources needed for acceptable AI performance.

To address developer ease of use and resource optimization, Intel has developed OpenVINO, an open-source toolkit that accelerates AI inference while offering a smaller footprint and a “write once, deploy anywhere” approach. It enables programmers to develop scalable and efficient AI solutions with relatively few lines of code. OpenVINO also can convert and optimize AI models to run them in data centers, in the cloud, or at the edge.

While OpenVINO makes AI more accessible for real-world scenarios, such as AI inference, it needs security and delivery features much like applications do. An OpenVINO deployment may consist of hundreds or thousands of container-based model servers. To reliably provide inference results to users or applications, you need a way to monitor server health and manage traffic. F5 NGINX Plus can secure OpenVINO model servers.

While it is possible to use the open-source version of NGINX with OpenVINO, using NGINX Plus adds valuable features like high availability, active health checks, and award-winning support. NGINX Plus also provides load balancing and SSL termination so apps and users can reliably and securely connect with your OpenVINO model servers.

For improved availability, NGINX Plus can be deployed as an active-passive pair. This creates a high availability solution, using the primary to actively process traffic while the backup monitors health and takes over in the event of an issue. Content caching and SSL termination improve performance by reducing the load on your OpenVINO model servers.

Active health checks in NGINX Plus monitor your model servers to ensure traffic is only sent to available servers. Secure and validate incoming encrypted traffic using mTLS certificates and use analytics and dashboards to anticipate or troubleshoot issues with your servers. The lightweight nature of NGINX Plus, as well as top-tier support for installation, deployment, and debugging, make it easy to adopt for your AI-powered application needs.