Traefik Labs has integrated its Triple Gate security architecture into HPE Private Cloud AI, a collaboration co-developed with NVIDIA, to enable organizations in regulated sectors to deploy sovereign AI infrastructure with complete offline capabilities and zero external dependencies.
The integration allows for the deployment of Traefik’s AI Gateway, MCP Gateway, and API Gateway on HPE Private Cloud AI, which is part of the NVIDIA AI Computing by HPE portfolio. This setup is designed to provide a sovereign AI runtime platform, addressing the needs of industries such as intelligence, healthcare, and financial services that require strict architectural control and data sovereignty.
The Triple Gate security architecture consists of three core components. The AI Gateway deploys NVIDIA Nemotron Safety models as NVIDIA NIM microservices on HPE infrastructure to filter malicious prompts and prevent jailbreak attempts, operating without external dependencies. The MCP Gateway enforces Task-Based Access Control for AI agents interacting with enterprise systems like Salesforce or Oracle databases, using declarative policies to define specific tasks and tools. The API Gateway safeguards backend services through centralized credential management, rate limiting, TLS/mTLS enforcement, and DLP inspection across all APIs utilized by agents.
Sudeep Goswami, CEO of Traefik Labs, stated, “The enterprise AI landscape is shifting from cloud-first to sovereignty-first, especially in regulated industries where architectural control isn’t optional—it’s existential.” Goswami added that the combination of Traefik’s Triple Gate security architecture with HPE Private Cloud AI and NVIDIA AI infrastructure provides the capability to run complete AI workloads entirely on customer-controlled infrastructure, maintaining consistent security and governance controls across cloud, on-premises, or air-gapped environments.
Key benefits of this integration include true sovereignty, allowing complete AI workloads to run on HPE infrastructure without external service dependencies, even in air-gapped environments. It also offers portable governance, enabling policies defined in Git to be deployed identically across various environments, and unified security, managing all three security layers through a single declarative platform.
Robin Braun, Vice President, AI Business Development for Hybrid Cloud at HPE, commented, “Through the HPE Unleash AI Partner Program, we’re bringing together trusted innovators like Traefik Labs to help enterprises confidently operationalize AI, securely.” Braun emphasized HPE’s focus on simplifying the process for companies to meet their security and governance objectives while leveraging AI’s potential.
As part of the HPE Unleash AI Partner Program, Traefik Labs has also joined the GreenLake Marketplace. This platform allows customers to discover software solutions compatible with HPE infrastructure and connect directly with partners to streamline software purchasing and license management.
Traefik Labs specializes in sovereign AI runtime infrastructure, integrating AI Gateway, MCP Gateway, and API Gateway into a unified declarative platform. With over 3.4 billion downloads, the company provides enterprise-grade security for cloud-native applications and AI workloads, partnering with NVIDIA, HPE, Oracle, SUSE, and Nutanix to deliver solutions for regulated industries requiring AI sovereignty.
