AI development projects are held back by a myriad of obstacles, including data sprawl (such as disparate data types and data stored in unconnected systems) as well as the complexity and cost of building and managing pipelines to connect data.
EnterpriseDB seeks to help enterprises overcome these barriers with its latest EDB Postgres AI platform update. This includes features such as Converged Analytics and Agentic Database, which create a single data foundation that customers can leverage to build agents and other AI applications.
Converged Analytics addresses the separation of operational and analytical data with an architecture that eliminates the need to build extract, load, and transform pipelines to connect the two, making data stored in EDB Postgres AI continuously available. Meanwhile, Agentic Database evolves EDB Postgres AI from a manually managed database to an autonomous database by using agents to monitor over 200 metrics to discover and resolve issues before they impact your workloads.
Additionally, the Agenttic database includes the ability to combine vector, JSON, and time series data through a SQL interface to improve the data retrieval process, making it easier for agents to access relevant data that provides the context they need to provide reliable output.
“We’re talking about integration here, not a single feature,” IDC analyst Devin Pratt told TechTarget. “Sprawl is the problem, and bringing relational, analytical, vector, and agent work into one managed Postgres foundation is the real solution to the problem.”
Matt Aslett, an analyst at ISG Software Research, similarly pointed to the value of EnterpriseDB in consolidating previously separate data workloads.
“Both Agentic Database and Converged Analytics are significant enhancements for existing and future EDB customers, reducing the cost and complexity of database management as well as architecture,” he said.
EnterpriseDB, based in Wilmington, Delaware, is a PostgreSQL database specialist with open source PostegreSQL platforms, database vendors such as MongoDB and MariaDB, and hyperscale cloud vendors offering PostgreSQL databases such as AWS, Google, Microsoft, and Oracle.
Integration for AI
Although success rates are improving, most enterprise AI projects still fail. While not necessarily the cause, data or data issues such as data clutter or poor data quality are often one of the main causes.
We’re talking about integration here, not a single feature. Sprawl is the problem, and bringing relational, analytical, vector, and agent work into one managed Postgres foundation is the real solution to the problem.
Devin PrattIDC Analyst
In response, many data management and analytics vendors this year introduced features aimed at helping customers organize, discover, and connect agents with data and business logic, giving agents the situational awareness to perform as intended. For example, in June alone, AWS announced a context layer for AI, Databricks similarly announced a context layer for agents as part of its Genie feature line, Microsoft added capabilities to turn the Fabric platform into the foundation for AI, and Snowflake announced tools to help manage and standardize agent context.
EnterpriseDB is currently adding similar features to aid AI development, partly motivated by customer feedback, according to Max Romanenko, the vendor’s chief engineering officer.
“Customer feedback has been consistent across the board: teams are spending a disproportionate amount of engineering time on data movement and database management, which should not require ongoing manual intervention,” he said.
Furthermore, market observation played a role in the development of EnterpriseDB’s Converged Analytics and Agentic Database, Romanenko continued, noting that Databricks and Snowflake each acquired a PostgreSQL database vendor in 2025 to add a data layer to their platforms.
“We’ve been in charge of that operational layer for 20 years, so the question for us was how do we surface what we already have so that we can deliver transactional, analytical, and now agent workloads as one integrated system,” he said.
To combine operational data that informs day-to-day operations with analytical data that informs long-term strategic decisions, Converged Analytics exposes operational data to Apache Iceberg sources, making it available to real-time and analytical engines, and queryable through a unified, managed PostgreSQL interface.
The results include significantly faster query speeds, data migration efforts reduced to hours instead of weeks or months, and lower cost of ownership based on a per-core pricing model where customers are billed based on the number of CPU cores used (with a fixed cost) rather than fluctuations in usage.
Similar to Converged Analytics, Agentic Database combines different data types. But beyond SQL interfaces for relational, JSON, time series, geospatial, and vector data, it’s designed to improve the operations of EnterpriseDB’s database platform with agents that proactively discover and resolve issues within your organization’s guidelines. Each customer can set up guardrails, including row-level and role-based access controls, and agents can enforce them, significantly reducing manual effort.
In addition to converged analytics and agentic databases, updates to EnterpriseDB include vector search, structured data, governance at the data layer that combines analytics, a bring-your-own-cloud (BYOC) option that lets customers apply AI to data where they reside, and EDB Developer Cloud, which provides a collaborative environment for AI development.
Converged Analytics and Agentic Database are generally available, while BYOC and EDB Developer Cloud are in preview and expected to be generally available in late 2026.
Aslett said Converged Analytics and Agentic Database are valuable additions, one that powers both real-time analytics and historical reporting, and another that automates database management and data management tasks, but they also help EnterpriseDB differentiate itself in a competitive market.
“Many software vendors offer PostgreSQL distributions and database-as-a-service, but EnterpriseDB is differentiated by its level of expertise and unique features,” he said. “These have traditionally included compatibility with Oracle and, more recently, advanced support for developing and deploying AI applications and agents both on-premises and in the cloud.”
Pratt said Converged Analytics is the most important new feature.
“Eliminating ETL between operational and analytical data is exactly what convergence companies are telling us is a priority right now,” he said.
Meanwhile, from a competitive perspective, the new features represent innovation and help differentiate EnterpriseDB from some of its competitors, Pratt continued, noting that Databricks is the vendor with the most similar capabilities to Lake Transactional/Analytical Processing and Lakebase..
“EnterpriseDB is pursuing the same convergence, but with open Postgres running on the customer’s own infrastructure,” he said. “For sovereignty-conscious companies, this is difficult to match.”
Appeal to consumers
According to Romanenko, just as user feedback helped drive the development of EntrepriseDB’s Agentic Database and Converged Analytics, conversations with customers informed BYOC and the governance features currently in preview.
“It felt like the right architectural extension of where we were already, applying rules associated with agents that were written to actually run within the data layer, rather than observing the agents from the outside,” he said.
Meanwhile, to appeal to potential new customers, Pratt suggested adding features to EnterpriseDB that would appeal to self-service developers. He noted that while EnterpriseDB has a strong base of large enterprise customers, its self-service capabilities could help it attract the smaller businesses that have helped vendors like Neon and Supabase grow.
“The natural next step is bottom-up implementation,” Pratt said. “EntrpriseDB is strong for large enterprises, and the self-service experience for developers will bring in the next generation of users along with that foundation.”
Eric Avidon is a senior news writer at Informa TechTarget and a journalist with more than 30 years of experience. He is responsible for analysis and data management.
