Aerospike raises $109 million to expand artificial intelligence real-time database

As the demand for artificial intelligence continues to grow, so does the interest and demand for database technologies that help support artificial intelligence efforts.

Aerospike is one such vendor that is seeing new demand for its real-time NoSQL database platform. To support this demand and help the company expand, the company today announced it has raised $109 million in a new round of funding as it seeks to add new AI capabilities and expand its go-to-market efforts. Aerospike was founded in 2009 and focuses on advertising technology applications. In recent years, Aerospike has been steadily adding new features and growing into a multi-model database, which will provide support for the JSON document model in 2022 and add graph database capabilities in 2023.

Aerospike is increasingly used to power artificial intelligence and machine learning (ML) applications, with its many users including Adobe, AppsFlyer, Barclays, Flipkart and PayPal. While Aerospike is currently used to support AI/ML, it does not yet have vector capabilities, which the company is building and plans to release this quarter.

“We were founded on the premise that real-time data access is going to be pervasive in every industry,” Aerospike CEO Subbu Iyer told VentureBeat. “One of the things we saw was the need to really get access to data and drive insights in real time.”

Vectors aren’t the only thing you need for an AI database

The last year has seen a surge in the number of database providers supporting vector capabilities.

Google recently announced that all of its cloud databases will support vectors. Graph database vendor Neo4j offers a competitive offering for Aerospike and announced vector support in August 2023. While vectors are critical for enabling different generative AI use cases, they are not the only way organizations can use databases to implement AI.

“Many of the use cases we do today fall into the realm of predictive AI,” Iyer said.

Aerospike currently supports AI and ML use cases without native vector support, Iyer said. These use cases often include predictive AI, a form of artificial intelligence used to make predictions that predates the deployment of modern AI-type models. For predictive models, organizations are training the models offline and then storing the results in Aerospike, Iyer said. Going forward, Iyer expects Aerospike to be used to support continuous learning processes.

Aerospike is currently building out full vector functionality as an enterprise preview for early access customers and will be generally available this quarter.

The intersection of graphs and vectors is a powerful AI combination

Aerospike’s existing multi-mode database capabilities, including a graph database, combined with the vector support under development, have some interesting possibilities.

“We see a lot of synergy between vectors and graphics,” Iyer said.

Iyer noted that organizations are leveraging vector embeddings to help bring more context and intelligence to AI. Some vectors can also be graphs, used to better understand the relationships between different entities.

“Our current graph solutions are primarily knowledge graphs, and we are selling identity and fraud,” Iyer said. “People are leveraging vectors for recommendation engines, fraud, etc., so there’s a natural overlap between vectors used to provide context and graphics that are part of the overall value chain.”

In addition to improving support for AI use cases, Aerospike plans to add more core database capabilities to its platform in the coming months. Planned improvements include enhanced multi-record transactions, point-in-time recovery, and improved observability and management capabilities, Iyer said.

“As our databases are actually being deployed to larger cluster sizes and large data sets, we want to make sure it’s easier to manage those clusters,” Iyer said.