MariaDB Acquires GridGain to Build Sub-Millisecond Data Infrastructure for Agentic AI
MariaDB will merge its relational database with GridGain's in-memory computing platform to target latency-sensitive agentic AI workloads, marking the company's third acquisition in nine months under private equity ownership.
Overview
MariaDB plc announced on March 9 that it has entered into a definitive agreement to acquire GridGain Systems, the company behind the open-source Apache Ignite in-memory computing platform, according to SiliconANGLE. Financial terms of the deal were not disclosed. The transaction, which remains subject to customary closing conditions, represents MariaDB’s third acquisition in nine months as the newly privatized database company repositions itself for the agentic AI era.
The combined platform will pair MariaDB’s ACID-compliant relational database with GridGain’s distributed in-memory data grid, targeting sub-millisecond latency for AI inference and autonomous agent workloads that traditional disk-based architectures struggle to serve.
What We Know
GridGain, founded in 2007 by Nikita Ivanov and Dmitry Setrakyan, is the original developer of Apache Ignite, an open-source in-memory computing platform that stores data as key-value pairs distributed across server clusters, as Blocks and Files reported. The company raised $52.1 million across three funding rounds, with its final $13.1 million Series C closing in 2021, and reported its tenth consecutive year of revenue growth with cash-positive operations as of 2023. In February 2025, GridGain launched a specialized product enabling low-latency data storage for real-time AI tasks including inference.
MariaDB CEO Rohit de Souza framed the acquisition around what he called the “AI latency gap.” “The rise of agentic workloads has placed unprecedented demands on enterprise infrastructure, causing requirements to explode and requiring a level of scale and submillisecond latency that traditional systems simply weren’t built to handle,” de Souza said in the announcement.
GridGain CTO Lalit Ahuja echoed the rationale: “Enterprises today cannot afford the latency introduced by siloed data architectures. The combined technology stack will unlock one of the key enablers for agentic enterprises: high performance and reliable data processing,” according to SiliconANGLE.
The merged entity will serve customers across financial services, telecommunications, and logistics, including American Express, Deutsche Bank, Nokia, Red Hat, Verizon, and UPS, per the announcement.
A Reinvention Under Private Equity
The GridGain deal is the latest chapter in MariaDB’s aggressive transformation since K1 Investment Management took the company private in September 2024 for $37.3 million, after a troubled stint as a public company. Under CEO de Souza, who was appointed at the time of the K1 acquisition, MariaDB has pursued a rapid acquisition strategy: it bought Galera Cluster from Codership in June 2025 for high-availability replication, then reacquired its former subsidiary SkySQL in August 2025 to regain a managed cloud database-as-a-service offering with serverless and agentic AI capabilities. In January 2025, the company launched MariaDB Enterprise Platform 2026 with built-in vector search for retrieval-augmented generation.
The pace stands in contrast to MariaDB’s position just two years ago, when the MySQL fork faced questions about its long-term viability. As previously reported, the broader MySQL ecosystem has been grappling with governance concerns, and MariaDB’s pivot toward AI-specific infrastructure represents a deliberate attempt to carve out differentiated territory rather than compete directly on the relational database feature treadmill.
What We Don’t Know
MariaDB has not disclosed the purchase price for GridGain, making it difficult to assess the financial scale of the deal relative to the company’s earlier acquisitions. The timeline for closing and for delivering an integrated product remains unspecified. It is also unclear how deeply the Apache Ignite in-memory layer will be embedded into MariaDB’s existing database engine, whether as a tightly coupled extension or as a separately deployable component within a unified platform.
The competitive landscape is also shifting rapidly. Snowflake’s $250 million acquisition of Crunchy Data, Databricks’ $1 billion purchase of Neon, and Supabase’s $5 billion valuation all reflect a broader consolidation wave in the database market driven by AI workload demands. Whether MariaDB’s smaller-scale, multi-acquisition strategy can compete with these well-capitalized rivals for enterprise AI infrastructure spending remains an open question.