Few technology companies manage to raise funds at a price level significantly higher than what they posted in 2021. This is the feat of the American company Databricks, specializing in data analysis. It has just raised more than $500 million, from several funds and one of the new global technology heavyweights, Nvidia. The financing round, led by the T. Rowe Price Associates fund, values the San Francisco-based company at $43 billion, compared to $38 billion at the end of 2021, when the company welcomed, among others, Microsoft, Google and Amazon. The reason for this success is quite simple: its profession is at the heart of the revolution triggered by generative artificial intelligence.
“Databricks allows large companies to apply artificial intelligence and deep learning on very large quantities of truly usable data,” explained its co-founder and CEO, Ali Ghodsi, to Le Figaro in 2021. Videos, emails , voice, medical imaging, satellite photos… All this rich material that companies accumulate needs to be worked on to achieve its full potential. Just like oil needs to be refined to produce its true added value. But this is precisely what the technologies developed by Databricks allow. This so-called “unstructured” data – as opposed to the tidy figures of a database – is precisely what will feed all the power and richness of a language model like ChatGPT. “Enterprise data is a gold mine for generative AI,” said Jensen Huang, founder and CEO of Nvidia. Databricks works with 10,000 clients around the world, including HSBC, AstraZeneca, Toyota, Shell and, in France, SNCF. “Sputnik effect”
Databricks was able to immediately see on its turnover the “Sputnik effect” that ChatGPT had in the emergence of generative AI among the priorities of companies. Between May and June, it recorded the best quarterly growth in its history, with a 50% jump in revenue, which exceeded $1.5 billion for the first time. To go much further in its value proposition in this ultra-high-growth niche, Databricks acquired the very promising start-up MosaicML this summer for $1.3 billion. Its software allows companies to create their large ChatGPT-type language models, from their own data, and at a very affordable cost, according to the company's promise. This marriage of know-how, combined with the powerful technological and financial partnership concluded with Nvidia, opens up even greater prospects for Databricks. “Every organization should be able to benefit from the AI revolution with more control over how its data is used,” insists Ali Ghodsi. We help them deploy tailored language models that will deliver a differentiated experience to their end users. »This new financing of $500 million delays the prospect of an upcoming IPO, which the company has been thinking about for several years, waiting for the most opportune moment. It allows it to continue to develop calmly, or even to make other strategic acquisitions in an ultra-competitive AI market, which requires having deep pockets. “Databricks is at the forefront of generative artificial intelligence for businesses,” said Alan Tu, investment analyst at T. Rowe Price