EDB integrates Postgres AI with NVIDIA to boost secure AI for government

EnterpriseDB has announced its inclusion in the NVIDIA AI Factory for Government reference design, providing a foundation for sovereign, secure, and scalable AI infrastructure targeted at government agencies and highly regulated enterprises.

The integration of EDB Postgres AI (EDB PG AI) with NVIDIA's reference design aims to enable organisations to deploy mission-critical artificial intelligence applications while maintaining security, assurance, and operational control over sensitive workloads. This follows research from EDB stating that 95% of enterprises plan to establish their own AI and data platforms within the next three years, with those highly committed to sovereignty achieving a return on investment up to five times greater.

AI sovereignty requirements

Nancy Hensley, Chief Product Officer at EDB, commented on the importance of sovereignty in the public sector and regulated industries:

"AI can't reach its full potential in government or regulated industries without sovereignty and choice. Our integration with NVIDIA gives agencies the flexibility to deploy EDB Postgres AI wherever their data and mission demand-on-prem, in the cloud, or in hybrid environments-while meeting high standards of security and compliance."

The EDB Postgres AI platform has been included in the NVIDIA AI Factory for Government reference design, a comprehensive blueprint intended to support demands for security, scalability, and regulatory compliance. The reference design is tailored for AI workloads in environments with strict regulatory and operational demands.

Platform architecture

EDB PG AI is structured to offer flexibility for sovereign AI and data deployments. Organisations can choose to implement the solution on-premises, in private or public cloud environments, or in hybrid configurations, depending on their operational and compliance needs. Within this architecture, the platform leverages NVIDIA NeMo Retriever and NVIDIA NIM microservices. These integrations are designed to facilitate a secure and high-performance pathway for developing, deploying, and operating AI models.

Key technical features highlighted include a "model-to-data" approach, which enables machine learning models, including those utilising NVIDIA NIM microservices, to be operated locally, ensuring data remains within sovereign boundaries. The system's integration with NVIDIA NeMo Retriever enhances the speed and effectiveness of embedding and retrieval processes, which are essential for Retrieval-Augmented Generation (RAG) AI systems.

Compliance and security

The announcement notes that EDB is pursuing FedRAMP Authorisation, a recognition intended for cloud products and services used by US federal agencies. EDB is using the Game Warden platform to accelerate this accreditation process. The company also states that its solutions are designed to support compliance with database Security Technical Implementation Guides (STIGs) and FIPS 140-3 standards, and deliver Iron Bank-hardened containers to federal customers.

EDB PG AI is presented as a unified platform that brings together data and AI layers, forming the basis of the NVIDIA AI Factory for Government reference design. By integrating with this framework, EDB aims to support public sector and regulated industry requirements for sovereignty, compliance, and operational flexibility.

Availability and future expectations

According to EDB, the capabilities of EDB Postgres AI make it suitable for organisations aiming to operationalise their data and deploy large language models (LLMs) within controlled, sovereign environments. The platform is based on the PostgreSQL database system and supports transactional, analytical, and AI workloads within a single environment. EDB claims support for high availability and hybrid management, and a global partner network supports the product.

Organisations already committed to sovereignty have reported up to 5× ROI, according to EDB's research, underscoring the emphasis on secure, localised control of AI data and workloads among enterprises and government entities.

EDB continues to invest in both technological development and community engagement through its participation in the broader PostgreSQL community. The company asserts that the integration with NVIDIA's framework provides further options for public sector and regulated customers to deploy AI infrastructure according to their specific security and compliance needs.