For many organizations, the shift to cloud computing has played out more realistically as a shift to hybrid architectures, where a company’s data is just as likely to reside in one of a number of clouds, as it might in an on-premise deployment, in a data warehouse, or in a data lake. Today, a startup that has built a more comprehensive way to assess, analyse and use that data is announcing funding as it looks to take on Snowflake, Amazon, Google and others in the area of enterprise data analytics.
Firebolt, which has redesigned the concept of a data warehouse to work more efficiently and at a lower cost, is today announcing that it has raised $37 million from Zeev Ventures, TLV Partners, Bessemer Venture Partners and Angular Ventures. It plans to use the funding to continue developing its product and bring on more customers.
The company is officially “launching” today but — as is the case with so many enterprise startups these days operating in stealth — it has been around for two years already building its platform and signing commercial deals. It now has some 12 large enterprise customers and is “really busy” with new business, said CEO is Eldad Farkash in an interview.
The funding may sound like a large amount for a company that has not really been out in the open, but part of the reason is because of the track record of the founders. Farkash, was one of the founders of SiSense, the successful business intelligence startup, and he has co-founded Firebolt with two others who were on SiSense’s founding team, Saar Bitner as COO and Ariel Yaroshevich as CTO.
At SiSense, these three were coming up against an issue: when you are dealing in terabytes of data, cloud data warehouses were straining to deliver good performance to power its analytics and other tools, and the only way to potentially continue to mitigate that was by piling on more cloud capacity.
Farkash is something of a technical savant and said that he decided to move on and build Firebolt to see if he could tackle this, which he described as a new, difficult, and “meaningful” problem. “The only thing I know how to do is build startups,” he joked.
In his opinion, while data warehousing has been a big breakthrough in how to handle the mass of data that companies now amass and want to use better, it has started to feel like a dated solution.
“Data warehouses are solving yesterday’s problem, which was, ‘How do I migrate to the cloud and deal with scale?’” he said, citing Google’s BigQuery, Amazon’s RedShift and Snowflake as fitting answers for that issue. “We see Firebolt as the new entrant in that space, with a new take on design on technology. We change the discussion from one of scale to one of speed and efficiency.”
The startup claims that its performance is up to 182 times faster than that of other data warehouses. It’s a SQL-based system that works on principles that Farkash said came out of academic research that had yet to be applied anywhere, around how to handle data in a lighter way, using new techniques in compression and how data is parsed. Data lakes in turn can be connected up with a wider data ecosystem, and what it translates to is a much smaller requirement for cloud capacity.
This is not just a problem at SiSense. With enterprise data continuing to grow exponentially, cloud analytics is growing with it, and is estimated by 2025 to be a $65 billion market, Firebolt estimates. Still, Farkash said the Firebolt concept was initially a challenging sell even to the engineers that it eventually hired to build out the business: it required building completely new warehouses from the ground up to run the platform, five of which exist today and will be augmented with more, on the back of this funding, he said.
“Firebolt created a SaaS product that changes the analytics experience over big data sets” Oren Zeev of Zeev Ventures said in a statement. “The pace of innovation in the big data space has lagged the explosion in data growth rendering most data warehousing solutions too slow, too expensive, or too complex to scale. Firebolt takes cloud data warehousing to the next level by offering the world’s most powerful analytical engine. This means companies can now analyze multi Terabyte / Petabyte data sets easily at significantly lower costs and provide a truly interactive user experience to their employees, customers or anyone who needs to access the data.”
