Hyperspace Develops Tailored Instances to Speed Up Database Queries

The surging interest in generative AI applications necessitates increasingly larger data repositories for things like model training information. These databases often require significant hardware resources and, depending on the management algorithms, may experience latency issues. Businesses frequently face hard choices regarding the balance of costs, efficiency, and data accuracy within their databases.

However, according to Ohad Levi, the CEO and co-founder of Hyperspace, this dilemma can be circumvented. Hyperspace is at the forefront of crafting specialized cloud services aimed at enhancing the efficiency of two key database operations: lexical and vector searches. Lexical searches focus on finding precise keyword matches within a database, while vector searches analyze the search query’s semantic significance and context.

Levi asserts that by utilizing a blend of FPGAs and GPUs, Hyperspace can offer search capabilities up to tenfold faster than what’s possible with conventional databases that lack such acceleration.

“Our service is designed to vastly improve large-scale data retrieval efforts, particularly for those in the AI and generative AI fields,” Levi shared with TechCrunch. “As unstructured data volumes continue to climb, surpassing traditional search methods, there’s a pressing need for solutions capable of addressing both lexical and vector search demands to keep pace with market requirements.”

Before establishing Hyperspace, Levi honed his skills as an optimization engineer at Intel and later as a product marketing leader at HP. His experiences with the limitations faced by conventional search technologies in major tech operations propelled him to join forces with former Intel design consultant Max Nigiri and create Hyperspace.

Rather than selling the instances outright, Hyperspace provides access to its managed database software, which currently operates on AWS. The databases support a mix of data types, including video, images, and text, with pricing set based on storage size and the volume of queries.

“Hyperspace represents a cloud-based managed database service, utilizing a software-as-a-service approach and pricing model based on usage,” Levi detailed. “Our experts craft tailor-made AI infrastructure solutions to address the intricate search challenges faced by businesses.”

Levi boasts that Hyperspace’s instances not only promise remarkable performance improvements but also boast a 5x increase in throughput and a 50% reduction in costs compared to standard databases. While Levi refrained from making direct comparisons with rivals, the potential cost and performance benefits raise the question: Can Hyperspace carve out a niche in a market dominated by established giants like Azure, AWS, and Google Cloud?

Levi is optimistic, citing early successes and partnerships with firms in the fraud prevention and e-commerce sectors such as Forter, Nsure, and Renovai. Moreover, Hyperspace has seen its annual recurring revenue and total contract volume triple over the past year.

Following a $9.5 million seed funding round led by Mizmaa, with contributions from JVP and toDay Ventures, Hyperspace plans to significantly expand its database service offerings and introduce a free starter plan.

Levi is confident about Hyperspace’s future, stating, “We are in a constant state of innovation, developing products that will propel the search industry forward while catering to the diverse needs of our clientele, from large enterprises to small and medium-sized businesses. With every passing day, as generative AI systems proliferate and the volume of data grows, the demand for more advanced search capabilities becomes increasingly clear.”