EnterpriseDB has introduced a suite of new features for EDB Postgres AI for WarehousePG, enhancing its capabilities with per-core pricing and real-time data streaming. These updates aim to assist users in managing the costs associated with AI development while ensuring that AI applications have access to timely data.
WarehousePG upgrade
WarehousePG, which debuted in April, is an open-source PostgreSQL data warehouse derived from the Greenplum Database project. It forms part of the broader EDB Postgres AI platform, launched in May 2024, which integrates data management, analytics, and AI functionalities.
PostgreSQL is celebrated for its adaptability, supporting a range of workloads including geospatial, JSON, time series, and vector databases. As a centralized repository, a PostgreSQL data warehouse efficiently channels data to analytics and AI tasks.
The latest enhancements to WarehousePG include:
- Real-time streaming data capabilities, facilitating immediate data access for AI applications.
- A predictable per-core pricing model, allowing users to manage AI development costs more effectively.
- Improved data observability and data sovereignty features, with flexible deployment options across various cloud environments and on-premises setups.
Industry analyst Carl Olofson, founder of DBMSGuru, emphasizes the significance of these updates. He notes that the introduction of real-time streaming and enhanced observability positions WarehousePG competitively against other leading commercial data warehouse solutions tailored for AI. Furthermore, the focus on data sovereignty addresses critical concerns regarding the international application of AI on diverse data sources.
EnterpriseDB, headquartered in Wilmington, Delaware, finds itself in competition with various database vendors, including MongoDB and MariaDB, as well as major cloud providers like AWS, Google, Microsoft, and Oracle that offer PostgreSQL databases.
As enterprises ramp up their investments in AI, spurred by advancements in technology since the launch of OpenAI’s ChatGPT in November 2022, the demand for efficient data management solutions has surged. Applications such as generative AI chatbots and AI agents are becoming essential for simplifying data exploration, enabling informed decision-making, and automating processes to enhance operational efficiency.
However, the development of these AI tools can be financially burdensome, primarily due to their reliance on vast quantities of high-quality data. The complexity of ingesting, integrating, and preparing data for AI applications requires significantly more computational resources compared to traditional analytics tools. This has led to increased expenditures for enterprises focused on AI development.
To alleviate these financial pressures, many vendors are prioritizing cost control measures. For instance, AWS has introduced data management features aimed at cost efficiency, while other database providers like Aerospike and Neo4j have implemented performance enhancements to help customers reduce their expenses.
EnterpriseDB’s per-core pricing model for WarehousePG addresses these concerns by charging customers based on the number of CPU cores utilized, thereby stabilizing costs. This approach contrasts with consumption-based pricing models, which can lead to unpredictable expenses based on usage. Kevin Petrie, an analyst at BARC U.S., highlights the importance of this pricing strategy, noting that software costs are a primary contributor to budget overruns in AI projects.
Additionally, the newly introduced data streaming capabilities are crucial for modern AI applications. Stream processing allows for the rapid transfer of data from its source to applications, which is vital for agentic AI workflows that rely on real-time information. Petrie notes the growing demand for data streaming, particularly for AI use cases such as fraud detection and price optimization.
Olofson reiterates the importance of streaming support in AI workflows, stating that it is essential for integrating AI processing beyond simple end-user queries.
Beyond the pricing model and streaming capabilities, the WarehousePG update encompasses:
- An AI-ready architecture featuring stream processing, native vector search and storage capabilities, and in-database machine learning with Python and MADlib.
- Flexible deployment options across any cloud or on-premises environment.
- Data governance features, including observability tools to monitor anomalies that could impact AI outputs.
Quais Taraki, EnterpriseDB’s chief technology officer, emphasizes that the WarehousePG update was driven by customer feedback and market research indicating a strong desire for unified data and AI solutions. With 95% of enterprises planning to integrate these elements within the next three years, the focus on data sovereignty was also a key consideration in the development of these features.
Looking ahead
With the latest updates to WarehousePG now in place, EnterpriseDB is setting its sights on enhancing the interoperability of the EDB Postgres AI platform in 2026. As database platforms evolve, the ability to seamlessly connect with external systems becomes increasingly important for organizations developing AI and analytics applications.
Taraki notes that the upcoming focus will be on improving interoperability within the AI and analytics ecosystem, which will enhance business continuity, security, and integration with open data and AI frameworks. Olofson supports this direction, suggesting that partnerships and integrations will further position EnterpriseDB’s offerings within the larger analytics and AI landscape.
As organizations seek comprehensive solutions in complex data management environments, the ability to collaborate with technology suppliers and consultants will be crucial for EnterpriseDB in delivering effective business solutions.
Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than 25 years of experience. He covers analytics and data management.
