At the bottom of the billions, Databricks has momentum and big plans
“HI, JUST CHECK can I put more? The bosses of promising startups are bombarded with such texts these days. Large funds in particular are rushing to seize a piece of the technological pie (see graph). Yet one founder seems to have received more than his fair share of pitches: Ali Ghodsi, CEO of Databricks. And he said yes to many. On August 31, the company confirmed that, just six months after a $ 1 billion financing deal, it had raised an additional $ 1.6 billion, valuing it at $ 38 billion, or $ 10 billion more. only after the previous cycle. Among Silicon Valley connoisseurs, these numbers solidify Databricks’ status as the most publicized company of the day.
Enjoy more audio and podcasts on ios Where Android.
The software manufacturer should soon be known further. Later this year, he is expected to hold the largest initial public offering ever (Initial Public Offering) of a software company, larger than that, at the end of 2020, of Snowflake, its most serious rival. Alternatively, some predict it could be snapped up by Microsoft as part of the biggest software acquisition ever. Whatever the outcome, there is substance to the hype. In the age of artificial intelligence, databricks could become (AI), what Oracle and its databases were once in the world of conventional enterprise software: the dominant platform on which applications are built and run.
Databricks was founded in 2013 to market Spark, an open source software that processes reams of data from different sources to form algorithms that then become the engines of AI applications. The company added features, including code that makes it easier for developers to program the system as well as manage their workflow, and offered the package as a cloud-based subscription service.
Yet Databricks only really took off when it added another component called “Lakehouse”. It is a combination of two kinds of databases, a “data warehouse” and a “data lake” (hence the coat rack). The two have historically been separated due to technical constraints and because they serve different purposes. Data warehouses are filled with well-defined corporate data that allows a business to examine its past, for example, the evolution of its sales, which is known as ‘business intelligence’ (BI). Data lakes are essentially a dumping ground for all kinds of data that can reveal the future of a business, including whether sales are likely to rise or fall. Yet this separation is increasingly inefficient and unnecessary, says Max Schireson of Battery Ventures, an investor in Databricks. “Doing BI and AI in different systems today is a bit stupid, ”he notes.
Businesses have jumped on what Databricks is offering, especially incumbents fearing they would be disrupted by a AIpiloted start. Comcast, a US broadband provider, uses it to allow its customers to use their voice to select movies; ABN Amro, a Dutch bank, to recommend services; and H&M, a fashion retailer, to optimize its supply chain. Databricks now claims more than 5,000 customers and annualized subscription revenue of $ 600 million, up 75% year-over-year.
Launch Databricks on Snowflake
Mr. Ghodsi has aimed even higher. “Ultimately all the data should be on Databricks,” he says. He plans to invest the newly raised capital to continue to grow and become the leader in Lakehouse systems. No one should blame Mr Ghodsi, who once taught computer science at the University of California at Berkeley, for his ambitions. However, achieving them will not be easy. Other companies are already establishing themselves in the region. It will probably be able to push back the three big cloud computing providers: Amazon Web Services, Google Cloud Platform and Microsoft Azure. Although they have more than enough resources to compete and provide AI packages, they share a big problem. Companies increasingly prefer not to store all of their data in a single cloud, fearing that they will get stuck with a single vendor. Instead, they opt for products, such as Databricks, that work across multiple clouds.
Snowflake is another story. She also builds houses by the lake. He also takes a different approach. As Databricks adds BI to his AI platform, Snowflake, which grew up in the world of data warehouses, adds AI to his cloud BI packaging, which means that their respective products will overlap more and more. While most of Databricks’ code is open source, Snowflake’s code is proprietary. And while Databricks primarily sticks to a “land and expand” strategy, where small software deals grow into larger ones, Snowflake practices a more conventional top-down sales model that focuses on big deals from the start.
All of this will be a battle over the next few years. But it could be abruptly cut short if Microsoft takes over Databricks. The software company is already one of Databricks’ investors and cooperates closely with it. Among other things, Azure offers a version of the Databricks platform and Microsoft uses its name in presentations on its strategy, which it rarely does with other companies. It would be a good fit. At its core, Microsoft is still a company that sells tools for developers to write apps and the platforms on which to run them. And Databricks is both a complement and a strategic threat: it allows data, rather than people, to write the code.
Data bricks ” Initial Public Offering isn’t meant to make the company public, some analysts say, but to put a price on it, so that negotiations can start somewhere. But the hype surrounding the company could thwart such plans. Snowflake is now worth around $ 90 billion. If Databricks’ Initial Public Offering surpasses that of Snowflake, its asking price may well be north of $ 100 billion. And like Pinterest, a social media company Microsoft considered buying earlier this year, it can get too expensive, even for a company as busy as the world’s largest software company. ■
For a more expert analysis of the biggest stories in economics, business and markets, sign up for Money Talks, our weekly newsletter.
This article appeared in the Business section of the print edition under the title “The Oracle of AI”