MariaDB buys back the company it sold two years ago

news
Aug 26, 20254 mins

Acquires SkySQL, returning its DBaaS platform to the product portfolio.

handshake, acquisition, happy, M&A, merger
Credit: Shutterstock

MariaDB, a company that as recently as last year was very much in transition due in part to the fact its new CEO, Rohit de Souza, is a specialist in preparing organizations for acquisition and resale, went in another direction Tuesday by making an acquisition of its own.

It brought SkySQL, the cloud-based database-as-a-service (DBaaS) platform which it spun off in late 2023 into an entity that was staffed by the former MariaDB employees who had built and supported the product, back into the fold.

In a release, the company said the move enables it โ€œto meet customer and market expectations for greater flexibility and deployment choice, including a range of self-managed and fully managed cloud offerings. MariaDB Cloud will encompass the SkySQL capability and form a key part ofย  the product portfolio.โ€

Michael Ni, VP and principal analyst at Constellation Research, said, โ€œfor database buyers, MariaDBโ€™s reacquisition of SkySQL and the addition of vectorDB support are less about dazzling innovation and more about reassuring customers that the company can finally deliver a coherent cloud story.โ€

MariaDB, he said, โ€œoffers buyers an open-source with hybrid deployment choice without the hyperscaler lock-in. The real question is whether this move stabilizes the roadmap or simply resets the clock on past missteps.โ€

Devin Pratt, research director, data management, at IDC, said that with the acquisition, MariaDB Cloud has evolved into a fully managed, streamlined service designed with agentic AI in mind.

The platform, he said, โ€œoffers serverless, pay-as-you-go pricing alongside provisioned single-node and high-availability clusters.โ€ He pointed out that it is available across AWS, Google Cloud, and Azure in more than 40 regions, and offers optional add-ons such as Backup, SkyAI Agents, and SkyDBA, โ€œgiving customers broad deployment flexibility and minimizing dependency on a single provider.โ€

This architecture, said Pratt, โ€œpositions MariaDB Cloud for an agentic AIโ€“driven future. Agents can be launched on demand and scale elastically. They support session-based state maintenance, and deployment across multiple regions enables adherence to data residency requirements. Multi-cloud support further boosts resilience and helps mitigate vendor lock-in.โ€

โ€œ[MariaDB Cloud] delivers serverless scalability, multi-cloud flexibility, and integrated AI capabilities, an essential combination for organizations building toward an agentic AI future,โ€ he noted.

Asked what prompted the re-acquisition and when negotiations first started, Vikas Mathur, MariaDBโ€™s chief product officer, said that when he and de Souza joined the company last year, โ€œour first task was talking with MariaDBโ€™s customers. We wanted to get a sense of what they loved and what new capabilities they were looking for us to add from a product perspective. It became clear quickly that customers wanted a fully managed cloud database service to support their journey to the cloud.โ€

Mathur added, โ€œwe looked at a few different options on how we could deliver this, and after thorough vetting our options, we found that SkySQL provided us with a solid foundation to be able to deliver DBaaS quickly and efficiently to our customers.โ€ย 

With the acquisition, MariaDB customers gain access to new serverless and agentic AI capabilities from SkySQL, he said.

Andy Neill, vice president of the technical counselor program at Info-Tech Research Group, said, โ€œ[the coverage] around the acquisition of SkySQL is focused on the ability to simplify MariaDB hosting with serverless compute. However, MariaDB is so easy to set up and manage that I donโ€™t think that is the most exciting point of this news.โ€

โ€œSkySQL provides an MCP server that allows creation of MariaDB instances in a serverless environment,โ€ he said. โ€œBy extension, any custom agentic AI can also connect to this MCP server and perform similar functions, allowing AI agents to create their own infrastructure when needed. Now we have to decide if giving AI agents this ability, which could make information more ubiquitous but at the same time incur more compute cost, needs to be strictly governed or freely enabled.โ€

MariaDB Cloud, said Pratt, โ€œdelivers serverless scalability, multi-cloud flexibility, and integrated AI capabilities, an essential combination for organizations building toward an agentic AI future.โ€

Financial details of the deal were not released.

Paul Barker is a freelance journalist whose work has appeared in a number of technology magazines and online, including IT World Canada, Channel Daily News, and Financial Post. He covers topics ranging from cybersecurity issues and the evolving world of edge computing to information management and artificial intelligence advances.

Paul was the founding editor of Dot Commerce Magazine, and held editorial leadership positions at Computing Canada and ComputerData Magazine. He earned a B.A. in Journalism from Ryerson University.

More from this author