MariaDB Launches AI Agent-Ready Database Platform

 MariaDB plc , the unified open database platform for enterprises running AI agents, announced MariaDB Enterprise Platform 2027 , a new release that delivers 56% higher peak throughput¹, greater resilience in multi-region deployments, and new tools to develop and run AI agents faster. What began as an open-source operational database now enables businesses to manage critical transactions, analyze data in real time, benefit from memory speed, utilize AI tools, and scale across multiple regions from a single platform. The new brand identity, introduced alongside this release, reflects the company’s direction: empowering businesses running AI with the world’s most trusted database platform.

Read more – Semgrep Named a Strong Performer in Independent Evaluation of Agentic Development Security Tools

Every AI agent running in production relies on the reliability of the data it receives. However, in most enterprises, this data is distributed across multiple operational databases, data warehouses, vector repositories, ETL engines, and caches. Each integration adds latency, synchronization delays, and a potential point of failure. And this complexity is set to increase. IDC forecasts that by 2027, the use of AI agents in the world’s 2,000 largest companies (G2000) will increase tenfold, while token payloads and API calls will increase a thousandfold.² It also estimates that companies without a high-quality, AI-ready data infrastructure will experience a 15% productivity loss by that year.³ Enterprise Platform 2027 is designed to meet this demand and support the applications that enterprises already use, enabling them to incorporate agents without adding another vulnerable layer to their technology infrastructure.

“The AI ​​race won’t be won by the smartest model, but by the companies that unify the infrastructure supporting their agents,” said Rohit de Souza, CEO of MariaDB plc. “Every system added to implement AI introduces another integration point, and those points fail. Agents need a solid, stable foundation that allows them to track every action they take. That’s why we’re extending capabilities from the transactional core that already underpins our customers’ most critical systems. MariaDB is no longer just a database. We’re the platform on which companies operating AI agents are built.”

Read more – SerpApi Earns Ethical Web Data Collection Trust Seal

What Enterprise Platform 2027 offers

While analytics-first vendors are trying to incorporate transactions, the most complex workload, MariaDB is expanding its capabilities from a transactional core that already underpins mission-critical systems across all industries, from financial services and telecoms to retail and the public sector. Each new feature thus benefits from that reliability. The platform is built on carefully selected open source code: MariaDB maintains the open source code while vetting contributors, evaluating releases, and distributing versions with hardened security. This gives businesses access to the transparency of open source and a responsible vendor. Every component of MariaDB Enterprise Platform 2027 builds on this foundation to meet the demands of modern applications and AI agents.

• Develop agents faster: AI Retrieval connects agents to real-time operational documents and data through a single managed service, converting natural language queries into read-only SQL queries. AI plugins bring MariaDB’s expertise to programming tools like Claude Code and Codex, enabling developers to generate error-free code the first time.
• Run agents faster: Agent Memory provides persistent in-memory storage, shared across sessions, so agents can pick up where they left off, even in long processes. Semantic Cache retrieves stored answers to questions already phrased differently, reducing response times and token costs.
• Unify data access: MariaDB Exa enables real-time analysis of operational data without complex data pipelines. Vector search, natively integrated into the engine, now processes up to 30% more queries per second than the previous major version, so AI applications get relevant results faster.
• Scale without rebuilding infrastructure: This version offers 56% higher peak performance than the previous version, based on third-party testing. Additionally, clustering via Raft and conflict detection and resolution features improve scalability and resilience in multi-region deployments.
• Ensuring security and control: A unified set of role-based access controls covers all workloads. Bring Your Own Cloud allows businesses to keep their data where they prefer, while FIPS 140-3 compliance ensures that federal agencies and other regulated entities can meet their cryptographic requirements.

Read more – Writesonic Traces What AI Says About a Brand to Its Source and Launches Outreach Agent to Fix It

“The database has become, almost without us realizing it, the most difficult problem to solve in enterprise AI,” said Devin Pratt, research director for database systems at IDC Enterprise Intelligence. “Most companies won’t rebuild their data infrastructure to adopt agents. They’ll incorporate them into the systems they already use to operate. That’s why the transactional core and its ability to deliver analytics, information retrieval, and memory from the same data in real time is the most practical starting point for agent-based applications.”

“Most platforms are moving toward agents from the analytics realm, which involves incorporating transactions, the most complex workload,” said Sanjeev Mohan, principal analyst at SanjMo. “Expanding capabilities from a proven transactional core is the most solid path, because agents need a stable foundation. The companies that succeed with agents will be those that eliminate data fragmentation and replace it with an infrastructure that agents can truly rely on.”

A new identity that reflects the evolution of MariaDB

Enterprise Platform 2027 is the result of two years of strategic investment and development, and its capabilities extend far beyond those of a traditional database. MariaDB’s new brand identity reflects this transformation, with a redesigned logo and a bolder visual system based on a vibrant color palette. The new identity is being launched today in conjunction with the revamped mariadb.com website . 

“A brand should convey a company’s convictions before anyone reads a single word, and ours was telling an outdated story,” said Tracy Strauss, Chief Marketing Officer of MariaDB plc. “Our new identity reflects a company that is open by design, unified across all workloads, and ready for businesses operating with AI agents. It’s vibrant and intentionally inspires confidence because that’s the company we’ve become.”