Edelbyte Ltd., the developer of a data analytics platform optimized for artificial intelligence agents, launched today with $70 million in funding.

The startup raised the capital over two rounds. The investments were led by Sequoia and Quiet Capital with contributions from Greenoaks and Swish Ventures. According to CTech, Edelbyte is valued at $500 million.

Tel Aviv-based Edelbyte was founded last year by former Microsoft Corp. executives Avner Aharoni, Evgeney Ryzhyk, Alexander Sloutsky and Yoni Leibowitz. They helped develop Azure Data Explorer, a tool for analyzing real-time information streams. Customers often refer to it as Kusto in a nod to the eponymous query language it uses to find data patterns.

The company positions its platform as an alternative to Snowflake and Databricks. It’s designed to store up to hundreds of petabytes of data per deployment. It can hold both standard records and vectors, the mathematical structures in which AI agents keep their knowledge. Vectors store not only data snippets such as product listings but also information that makes it easier to find similar items.

The platform supports a wide range of queries. A search engine can use it to find records sought by users, while a business intelligence tool can run analytics calculations. There’s also support for graph processing. That’s an analytics approach used to study data points that relate to one another in some way, for example because they were created in the same system.

Edelbyte claims that its platform can query multipetabyte datasets with under 100 milliseconds of latency. Under the hood, it uses several different optimization methods to achieve that performance .

Analyzing a dataset usually requires companies to move it to a dedicated processing environment. Edelbyte uses a technology called query federation to skip that step. Instead of requiring users to load all their data into its platform, It can run queries in the systems where the data is usually stored. It thereby speeds up processing and reduces the amount of bandwidth that must be spent on moving data between environments.

Edelbyte further optimizes queries by generating indexes. Those are collections of shortcuts that make it easier to find specific items in a large dataset. An index only works if it’s tailored to the data access patterns of the application that uses it. According to Edelbyte, its platform automates some of the manual work involved in the index customization workflow.

“The systems companies rely on to search and analyze their data were built for a world before AI,” said Chief Executive Avner Aharoni (pictured, right). “As it becomes central to how businesses operate, the volume and complexity of data queries will grow far beyond what legacy infrastructure was designed to handle. We believe this creates an opportunity to build a new market leader in search and analytics.”

Edelbyte says several Fortune 500 customers use its platform to power petabyte-scale data environments. The company will use its new funding to grow its installed base and hire more employees.

Photo: Edelbyte