MariaDB spinout SkySQL secures seed funding to ‘bring conversational AI to databases’



Anyone who’s followed the fortunes of MySQL and MariaDB these past 15 years will probably remember SkySQL, another brand that once existed within that same database ecosystem and, a year ago, became a standalone company once more.

That company today announced it has raised its first outside funding — a $6.6 million seed round — with ambitions to use it to bring conversational AI to databases.

MySQL, for the uninitiated, is the popular relational database management system (RDMS) created by Michael “Monty” Widenius, David Axmark, and Allan Larsson back in the mid-’90s. After the commercial company behind the project was sold to Sun Microsystems for $1 billion in 2008, Sun Microsystems itself was snapped up by Oracle for $7 billion, a transaction that spurred Widenius to fork MySQL and create MariaDB to ensure it remained an independent, community-driven effort.

Subsequently, Widenius established two organizations: SkySQL, a commercial entity that sold services for MySQL; and Monty Program, which was all about supporting MariaDB.

These two businesses merged in 2013, with the combined outfit rebranding as MariaDB Corporation and raising some $230 million in funding. MariaDB resurrected the SkySQL brand in 2020 via a new cloud database product, but after struggling with life as a public company (read more about that saga here), MariaDB spun out SkySQL as an independent business last year, spearheaded by the team that had been developing the product at MariaDB.

While MariaDB retained an equity stake in the spin-off, the onus has very much been on SkySQL to forge its own path. And that is why the startup today announced a fresh cash injection, as it looks to bring AI into database mix.

The startup’s $6.6 million seed funding includes participation from renowned seed-stage investor Eniac Ventures, which has previously invested in the likes of Airbnb and Reddit, in addition to Good Capital and WTI.

The AI factor

SkySQL Platfrom
SkySQL PlatfromImage Credits:SkySQL

Businesses typically use SkySQL to alleviate the hassles and complexities of running MariaDB — it’s a fully-managed cloud product that automates provisioning and scaling, with features dealing with security, compliance, and disaster recovery. And since spinning out of MariaDB last year, SkySQL has added Microsoft Azure support, meaning it’s now truly multi-cloud as it already supported deployments across AWS and Google Cloud Platform.

However, the company is now looking to align itself with the burgeoning AI revolution, and tackle one of the biggest problems it says exists in today’s databases — that is, they’re not optimized for AI.

As such, SkySQL is now introducing a duo of AI products, including something it’s calling “semantic agents” that serve insights on operational data. So, SkySQL customers can effectively build and embed their own AI agents into their apps, allowing their users to ask natural language questions of their data stored in SkySQL — something that can be challenging in enterprise settings due to security and compliance issues related to moving sensitive data around, not to mention longstanding issues around so-called “hallucinations.”

SkySQL keeps everything within the platform, while the semi-autonomous agents also support human-in-the-loop input to help refine the output and improve it over time.

Creating a semantic agent in SkySQL
Creating a semantic agent in SkySQLImage Credits:SkySQL

Elsewhere, SkySQL is also debuting SkyCopilot, which is basically an AI assistant that can help database administrators optimize the performance of their database, troubleshoot, and “manage routine tasks” all through natural language queries. This could be something like “please give me a report summarizing the health of my server” or “identify slow-performing database queries.”

So in effect, SkySQL is positioning itself as much more than a database company, it’s all about unifying various integral AI components, such as vector databases and large language models (LLMs), into a single API that allows users to derive natural-language answers without having to move their data.

Founding team

SkySQL’s founding team includes CEO Nithin Rao (pictured above), who co-founded a vehicle connectivity platform called Autonomic which he sold to Ford back in 2021. He then launched an AI analytics startup called Definitive Intelligence, which was snapped up by AI chip company Groq this past March, though Rao had already departed for SkySQL by that point.

“SkySQL represented a next natural step — a chance to deliver on the vision of being the operational data platform for modern applications, especially those leveraging generative AI,” Rao told TechCrunch.

Rao’s two co-founders are CTO Jags Ramnarayan and CPO Saravana Krishnamurthy, both of whom transitioned over from MariaDB, where they were already familiar with SkySQL. And although Rao only joined them as SkySQL was emerging from MariaDB, he already knew one of the founders from a previous cloud computing company they worked at some years previous.

“Jags and I first crossed paths while working together at Pivotal, and when the opportunity to spin out SkySQL arose, we reconnected,” Rao said. “The opportunity to join forces with such a strong team and build something transformative in this space was one I couldn’t pass up.”




Source