Upsolver provides software that eliminates the high engineering overhead of operating cloud data lakes, raised $13 million in Series A financing. Vertex Ventures US led the round with participation from new investor Wing Venture Capital (Wing) and existing investor Jerusalem Venture Partners (JVP).
Data Lake projects are known for being lengthy, code intensive and complex. There is no self-service for data consumers and data engineers, get bogged down with endless ETL work with Hadoop or Spark. Upsolver reduces high-maintenance ETL issues as it uses familiar SQL interface, allowing analysts and engineers to join semi-structured streams and historical big data into actionable analytics and machine learning.
Upsolver is built to run natively and securely on an AWS account by decoupling storage on S3, compute on EC2 and metadata management in Glue Data catalog.
The investment signals a growing need for Upsolver’s cloud-native solution. Upsolver has tripled overall company ARR and seen increased demand during the pandemic, having low churn, and significant expansions amongst customers.
Modern enterprises use cloud data lakes to analyze large volumes of structured and unstructured data by breaking the traditional database into three pieces: storage, compute, and metadata. This separation dramatically reduces both cost and dependency on one database vendor but it introduces a new engineering complexity — each piece must get configured, optimized, and synchronized with the rest. This time-consuming and cost-prohibitive process could historically only be completed by big data engineers who code and operate open-source software like Apache Spark or Hadoop.
“Big data engineers are a unicorn hire,” said Ori Rafael, Upsolver’s CEO and co-founder. “They should spend their time solving an organization’s hardest data problems instead of performing repetitive tasks like job orchestration, ETLs, and IT management. Upsolver helps automate repetitive tasks with a powerful tool that can be used by existing data practitioners. Our customers see an average of 95% reduction in the data lake management effort.”
In Sik Rhee, General Partner and co-founder at Vertex Ventures US, who founded Opsware (acquired by HP for $1.6 billion) and made early-stage investments in both Cloudera and Couchbase, will join Upsolver’s board of directors and so will Gadi Porat, Partner at JVP. Additional investors include Wing, which made an early-stage investment in Snowflake, Jeff Rothschild, founder of Veritas and First Senior Technology Executive at Facebook, and Sohaib Abbasi, former CEO and Chairman of Informatica.
“We see Upsolver creating a cloud-native standard for data lake computing,” said Rhee. “Upsolver succeeded in abstracting away the engineering complexity of data pipeline management so that enterprise customers can quickly solve their modern data challenges in real-time and at any scale, without having to build another silo of expertise within the organization.”
Upsolver founders Ori Rafael and Yoni Eini first met in Israeli intelligence, where Rafael was Head of Data Integration Platforms, and Eini was CTO of a large data science group. They came up with the idea for Upsolver while struggling with a data lake they built for advertising optimization. Being their users from the start has led to their product’s success.
Upsolver’s status as an Amazon Web Services (AWS) Partner Network (APN) Advanced Technology Partner also fueled its momentum. Through the APN program, Upsolver’s platform gets deployed and bought via the AWS marketplace. The Upsolver platform natively plugs into services like AWS Athena and Redshift, making it easy to set up a data lake on AWS.
Upsolver has raised a total of $17 million to date, with a global team working across Israel, California, and New York. Its customers include Asurion, Cox Automotive, IronSource and Sisense. Upsolver will use the funds to expand its R&D and Sales teams and enhance its multi-cloud capabilities.