Hybrid Manager 1.2 release notes

Release Date: June 17, 2025

We are thrilled to announce the first public availability release of EDB Postgres AI Hybrid Manager v1.2.

EDB Postgres AI Hybrid Manager is the control plane of the EDB Postgres AI Platform—a sovereign AI, analytics, and Postgres management solution. This release revolutionizes how you work with data by unifying advanced analytics, generative AI, and operational management into a single, cohesive platform.

Release Highlights

• Sovereign AI & Analytics Platform: Hybrid Manager is now a complete platform for building modern data and AI applications. Deploy a Postgres-native lakehouse and sovereign GenAI workloads with unified governance, all within your own infrastructure.
• New Deployment Option: Sovereign AI and Data Factory: Introducing a fully integrated hardware and software appliance, co-designed by EDB and Supermicro®, for deploying production-grade Postgres and AI workloads in your own data center with no external dependencies.
• Postgres-Native Lakehouse: A groundbreaking analytics stack is now built-in. Deploy Lakehouse Clusters to run high-speed, vectorized queries directly against open data formats like Apache Iceberg and the Delta Lake format in your object storage.
• AI Factory for GenAI Workloads: Build and serve powerful AI applications with a low-code GenAI Builder, integrated model serving with GPU acceleration, in-database Semantic Search, and a comprehensive Python SDK.
• Unified Management & Proactive Diagnostics: Experience a completely redesigned monitoring interface, a new cluster-level scorecard, proactive query recommendations, and deep query latency statistics.

Deployment Options

EDB Postgres AI Hybrid Manager v1.2 is available in two ways, providing maximum flexibility for your operational needs.

Sovereign AI and Data Factory (Hardware Appliance)

For organizations requiring maximum control and zero external dependencies, the Sovereign AI and Data Factory is a fully integrated, pre-assembled hardware and software platform. Delivered as a rack-mounted system to your data center, it is the ideal choice for environments where data sovereignty, security, and lifecycle reliability are paramount.

This engineered system is optimized for scale and is ready to run advanced Postgres, vector, and AI pipelines the moment it's installed. It includes validated Supermicro servers, lifecycle-managed Kubernetes, and the full Hybrid Manager software suite with optional GPU nodes for AI inference.

Software-Only Deployment

For organizations with existing infrastructure, EDB Postgres AI Hybrid Manager can be deployed as a software-only solution on your preferred Kubernetes distribution. It is certified for Amazon EKS, Google GKE, and Red Hat OpenShift (RHOS).

New Features & Enhancements

Sovereign AI & Analytics Platform

Hybrid Manager v1.2 introduces a powerful, integrated suite of tools for building and managing next-generation analytics and AI workloads directly on your Postgres platform.

Postgres-Native Analytics

Build a modern, cost-effective data architecture that combines the power of Postgres with the scale of object storage. Hybrid Manager provides the tools to manage your entire analytics stack.

• Lakehouse Clusters: Provision and manage dedicated compute clusters that provide scalable, vectorized query execution on object storage. They serve as the query engine for your Postgres-native lakehouse.
• Apache Iceberg Support: Natively manage large analytical tables in Iceberg format. Use Tiered Tables with EDB Postgres Distributed (PGD) to automatically offload cold data partitions to Iceberg, making them queryable via Lakehouse Clusters while reducing storage costs.
• Delta Lake Format Integration: Seamlessly integrate with existing data lakes. Lakehouse Clusters can directly query tables using the Delta Lake open format, enabling fast analytics on data produced by other big data pipelines like Spark.
• Tiered Tables with PGD: Efficiently manage large time-series datasets by automatically partitioning tables and offloading cold partitions to object storage in Iceberg format. A single query can span both hot transactional data in Postgres and cold historical data in your data lake.

GenAI Workloads with AI Factory

EDB Postgres AI Hybrid Manager v1.2 introduces the AI Factory, a comprehensive GenAI capability stack designed for enterprise-grade deployments.

• GenAI Builder: A web-based low-code interface to rapidly prototype, iterate, and deploy AI applications—from simple one-shot prompts to complex RAG pipelines.
• Model Serving: Deploy and manage NVIDIA NIM containers for AI inferencing on GPU-accelerated infrastructure, with built-in support for availability and autoscaling.
• Semantic Search: Use vector search capabilities built into Postgres to power retrieval for your GenAI applications. Ingest and embed text, HTML, PDFs, or images—then query them via similarity search.
• Python SDK for AI Development: The EDB AI Builder Framework SDK enables developers to assemble AI workflows using Python. Integrate retrieval, reranking, embeddings, and generative responses with fewer lines of code.

Observability, Monitoring & Insights

v1.2 enhances the platform’s observability with expanded metrics, sustainability insights, and performance diagnostics.

• Redesigned Monitoring Dashboard: Gain deeper insight into cluster health with a revamped UI, improved layout, and more granular control over metrics and thresholds.
• Query Latency Metrics: Visualize query latency statistics collected via the pg_stat_monitor extension directly in the Query Diagnostics tab.
• Proactive Query Recommendations: Receive optimization tips, including new index suggestions and stale statistics alerts, based on real workload patterns.
• Cluster Scorecard: View a unified summary of diagnostics, recommendations, and risk indicators for each cluster in one place.
• Self-Managed Postgres Support: Connect external Postgres instances to Hybrid Manager to collect query diagnostics, recommendations, and monitoring metrics.

Sustainability: ESG Reporting (Preview)

EDB Hybrid Manager now includes experimental support for ESG dashboards, helping organizations track energy usage and carbon emissions for their data infrastructure.

• Carbon and Energy Metrics: Visualize CPU-based energy usage and estimated CO₂ emissions using configurable emission factors.
• Region-Aware Reporting: Customize dashboards with local energy cost ($/kWh) and emission rate (kg CO₂/kWh) based on the deployment region.
• Integrated Grafana Dashboards: Access the new Power Usage & Carbon Footprint dashboards directly from the Hybrid Manager UI.
• Preview Phase Goals: Validate demand for ESG observability and collect feedback from interested users.

Platform Improvements

Resources

• Installing Hybrid Manager
• Using Hybrid Manager
• AI Factory Overview
• Analytics Lakehouse Architecture
• Power & ESG Reporting
• Release Notes Archive

Released: 17 June 2025