Ali Ghodsi, co-founder and CEO of Databricks Inc., speaks during a televised interview with Bloomberg Technology in San Francisco on Oct. 22, 2019.
David Paul Morris | Bloomberg | Getty Images
Databricks said Wednesday that its software for storing and analyzing large data sets will be available in Google’s public cloud.
The launch opens a new growth channel for a company that wants to be ready for an IPO this year. At the same time, Google is gaining access to a trendy piece of software previously only available through the two largest US cloud rivals, Amazon and Microsoft.
Databricks has also fine-tuned its software to enable customers to move systems running in Google’s cloud to their on-premises servers with minimal effort, co-founder and CEO Ali Ghodsi said in an interview with CNBC on Tuesday. Historically, Databricks was only available as a cloud-based service.
“We’ve been forced to do this by customers for years,” he said.
To get there, Databricks worked with Google engineers to modify the software to run in a virtual container, a lightweight alternative to the more traditional virtual machine software developed by VMware. The software runs on top of a Google cloud service based on Kubernetes, a container management tool that Google released under an open source license in 2014.
Kubernetes abstracts the underlying computing infrastructure, freeing programmers to do more innovative work, said Google Cloud CEO Thomas Kurian.
The arrival of Databricks in Google’s cloud market could lead to an increase in the consumption of computing and storage resources in the Google cloud, lagging behind Amazon Web Services and Microsoft Azure. Snowflake, whose software stores a variety of data, added support for Google’s cloud in 2019, ahead of its 2020 IPO.
“We are generating billions in revenue to the cloud providers, so they need us,” Ghodsi said in an interview earlier this month. “We are the killer app.” Organizations rely on Databricks to process many different types of data so that it can be used in applications or explored in data analysis tools such as Looker, owned by Google.
Databricks and Google first started talking about a collaboration two years ago. Ghodsi said he spoke with Kurian the week he joined Google, replacing former VMware CEO Diane Greene.
“Google transformed after it joined,” Ghodsi said. “We’re even seeing the same team members we worked with before, just the type of cadence and execution ability has changed dramatically, and it’s great to see.”
CapitalG, a corporate venture division within Google parent Alphabet, took part in the most recent Databricks funding round, announced earlier this month. Databricks had been in talks with CapitalG for over a year, Ghodsi said.
Databricks exceeded $ 425 million in annualized recurring revenue in fiscal 2021, up 75% year-on-year, Ghodsi said. According to LinkedIn, the company has more than 1,800 employees.
WATCH: Thomas Kurian from Google Cloud on the future of cloud computing