MariaDB goes agentic and gets accelerated analytics boost

The MariaDB Enterprise Platform now has RAG pipeline and AI agent support, and an Exasol-based Exa in-memory engine accelerates analytics.

The company provides the open-source MariaDB relational database, available as a free community edition (MariaDB Community Server), An enterprise edition either paid-for support (MariaDB Enterprise Server) and MariaDB Cloud, a fully-managed database-as-a-service offering. Native vector storage and search support was added in January this year. The MariaDB Enterprise Platform 2026 release unifies transactional, analytical and AI (vector) database engines into a single, high performance platform supporting retrieval-augmented generation (RAG) pipelines and AI agents with MCP support.

Chief product officer Vikas Mathur said: “The future of applications is agentic. AI agents need to probe, analyze and transact in real time and at enormous scale. At the same time, agents need to be grounded in insights contained in enterprise data that is trapped in fragmented silos today. MariaDB Enterprise Platform 2026 is purpose-built to eliminate that architectural friction.”

The company says its built-in RAG provides a RAG-in a-Box capability. Gen AI Large Language Models (LLMs) can use this to get context from the data stored in MariaDB, “speeding the development of GenAI and Agentic applications.” This eliminates the need for external embeddings, vector stores or retrieval pipelines. The company says itsRAG capability ensures all the LLM response generation steps are followed automatically, consistently and fully optimized.

There are ready-to-use copilots, agents, inside the database that enable users to interact with the database using natural language. Pre-configured agents include a developer and a DBA AI copilot and are available via MariaDB Cloud. The developer copilot is a Text-to-SQL agent connected to MariaDB’s database that responds to natural language queries by applying SQL queries to the database and returning the results. The DBA copilot executes DBA tasks like performance tuning or debugging errors.

Integrated MCP Servers enable AI Agents to “seamlessly interact with MariaDB databases and other databases across the enterprise.” They can launch serverless databases in the cloud and connect directly with MariaDB AI copilots.

The company says agentic application development platforms rely on serverless databases to meet the demands of AI agents, and “MariaDB Cloud offers a serverless database that provides the elastic scalability, operational simplicity and cost efficiency, which provisioned, always-on setups cannot match.”

In initial benchmarking, MariaDB Enterprise Server 11.8, the core server at the heart of MariaDB Enterprise Platform 2026, achieved 250 percent better performance compared to the 10.6 release.

A new MariaDB enterprise manager tool offers observability and management features, with topology-aware monitoring and a suite of visual tools for query development and schema management.

There is an enhanced database firewall in the MaxScale component of the MariaDB Enterprise Platform. This can programmatically record and enforce rules over how users query data.

All MariaDB customers have access to download the latest release immediately.

MariaDB Exa

MariaDB Enterprise Platform 2026 introduces MariaDB Exa designed for multi-terabyte, complex analytics on operational data at speeds well over 1,000 times faster than traditional OLTP engines and many times faster than leading analytical engines. MariaDB Exa integrates Exasol’s analytics engine and there is a strategic partnership between MariaDB and Exasol, with Exasol claiming it makes the world’s fastest analytics engine.

MariaDB Exa, combined with MaxScale, uses a multi-node, high availability, architecture. Exa has an “in-memory engine designed for near real-time analytics, data warehousing, and AI/ML workloads. It leverages a massively parallel processing (MPP) architecture to execute complex queries with unmatched speed and price/performance ratio.” Customers can “run AI/ML models directly within Exasol to eliminate data movement, reduce latency, and accelerate model execution.”

MariaDB CEO Rohit de Souza said: “MariaDB Exa allows organizations to extract real-time insights from large operational datasets very effectively and get immediate value from transactional information without the complexity of moving the data to a separate analytics system.”

Exasol CEO Joerg Tewes added: “MariaDB customers can now deploy a truly converged database solution that supports the performance requirements of GenAI, real-time dashboards, and complex data science – all while maintaining data governance and strict cost control.”

The MariaDB Exa solution is available immediately as an add-on option to the MariaDB Enterprise Platform 2026 release, with plans to incorporate the offering into MariaDB Cloud.