JFrog unveils new runtime security and Nvidia integration for AI model protection

Software supply chain company JFrog Ltd. today announced a new runtime security solution and a new product integration with Nvidia Corp. that provides users with enhanced security and the ability to secure artificial intelligence models.

JFrog’s new JFrog Runtime offers end-to-end protection for applications throughout their entire lifecycle, from development to deployment and production. The new service integrates into DevSecOps workflows, allowing organizations to implement security measures at every step of the software supply chain. In doing so, the service tackles vulnerabilities in real time to ensure that cloud-native applications, such as containers in Kubernetes environments, are monitored for potential risks.

Features of JFrog Runtime include real-time vulnerability detection and risk prioritization that allows security and development teams to identify security issues based on their business impact and hence accelerating the triage process. The platform also safeguards applications against potential post-deployment threats, such as malware or privilege escalation attacks, through advanced monitoring for cloud-based workloads.

JFrog Runtime also enhances collaboration between security and development teams through the provision of a unified platform for managing risks. With the unified platform, developers can track software packages from various sources to ensure compliance and version control, while at the same time, security teams can enforce policies that maintain the integrity of the software throughout its lifecycle.

Announced alongside JFrog Runtime today was JFrog partnering with Nvidia to integrate Nvidia NIM microservices into the JFrog Platform, enabling enterprises to deploy secure, graphics processing unit-optimized AI models quickly. Nvidia NIM is a microservices platform within Nvidia AI Enterprise that provides GPU-optimized infrastructure for deploying high-performance AI models and large language models securely and efficiently.

The new integration allows organizations to leverage Nvidia’s AI infrastructure for high-performance machine learning while maintaining visibility and security through JFrog’s unified DevSecOps workflows.

Through the combination of JFrog Artifactory with Nvidia NIM, JFrog says that enterprises can streamline AI model management and accelerate the deployment of LLMs. The platform also offers users centralized control that ensures compliance and traceability across AI deployments, from development to production.

“As enterprises scale their generative AI deployments, a central repository can help them rapidly select and deploy models that are approved for development,” Pat Lee, vice president of Enterprise Strategic Partnerships at Nvidia, commented on the new integration. “The integration of Nvidia NIM microservices into the JFrog Platform can help developers quickly get fully compliant, performance-optimized models quickly running in production.”

JFrog Artifactory provides a single solution for housing and managing all artifacts, binaries, packages, files, containers and components for use throughout software supply chains. The JFrog Platform’s integration with Nvidia NIM is expected to incorporate containerized AI models as software packages into existing software development workflows.

Photo: JFrog