Zilliz raises $60M and moves to SF

In 2020, Chinese startup Zilliz, which creates cloud-based native data processing software for artificial intelligence and unstructured data analytics applications and is the creator of Milvus, a popular open-source vector database for similarity search, raised $43 million in funding to scale the business and prepare the company to enter the U.S. market.Nearly two years later, Zilliz announced an additional $60 million in funding in conjunction with its eventual move to the West and the opening of a new

The round, led by Prosperity7 Ventures, a $1 billion venture fund from Saudi oil giant Aramco (the name refers to the first commercial well to produce oil), also includes previous Chinese backers Temasek's Pavilion Capital, Hillhouse Capital, 5Y Capital and Yunqi Capital. The company is not disclosing its valuation, but it's worth noting that this latest infusion is described as a $43 million Series B continuation rather than a new round. We'll update this information if we learn more. The total amount raised by the company is $113 million.

The capital and the move speaks not only to a key moment for the company, but also for the field of machine learning and broader trends affecting startups based in China.

The first of these highlights the explosive growth in popularity of Zilliz's open-source Milvus product. The company said the number of downloads has surpassed 1 million, up from 300,000 a year ago, and the number of users has grown 300% over the same period, but did not disclose the number of active users. Its customers include companies such as eBay, Tencent, Walmart, Ikea, Intuit and Compass.

As we noted back in 2020, Milvus doesn't resort to advertising and marketing spend, preferring to use word of mouth to attract attention in places where developers like to connect for ideas and inspiration, such as GitHub and Reddit. This strategy has paid off: the number of "stars" on GitHub has grown 200% to over 11,000, and the number of authors has doubled. (By comparison, it has been tagged about 4,400 times in 2020).

The reason for the interest in Milvus, as well as Zilliz's roadmap based on building new products, particularly the Zilliz Cloud service, which is currently in closed preview, and Towhee, another open source framework for ETL vector data, lies in the growing interest in vector databases and their use in artificial intelligence applications.

Simply put, while data can be (and often is) processed through more traditional databases, the complexity and structure of activities such as anomaly detection, recommendations, ratings, and other AI-driven tasks lend themselves more naturally and efficiently to vector databases built to handle the way AI data is represented. (Zilliz notes that its vector database "is cloud-based and can process billion-scale vector data in milliseconds.")

"Zilliz's journey to this point began with the creation of Milvus, an open source vector database that later became part of the LF AI & Data Foundation as a top-tier project," Charles Xie, founder and CEO of Zilliz, said in a statement. "Milvus has now become the world's most popular open source vector database with over a thousand end users. We will continue to be a major contributor and committer to Milvus and deliver on our promise to provide a fully managed vector database service in the public cloud that delivers the security, reliability, ease of use and availability that enterprises demand."

Other companies (Zilliz's competitors) like Pinecone and Weaviate are also developing solutions to address this problem. Pinecone raised money earlier this year and is backed by some impressive names, including Tiger Global and Menlo Ventures (for comparison, Pinecone has a valuation of $168 million, according to PitchBook). Weaviate's parent company, SeMI Technologies from the Netherlands, has also raised money this year, backed by funds such as NEA.

Meanwhile, major cloud providers such as AWS also have their own solutions. All of this points to a market opportunity that Zilliz is eager to capitalize on.

Notably, back in 2020, Zilliz said that more than half of Milvus users were outside of China: this speaks to how the company has positioned itself over time and how it sees its growth in the long term. Xi, who goes by the nickname "Starlord" (yes) and previously worked as a software engineer at Oracle in the US before moving to Shanghai to found Zilliz, told TC that from day one he thought of the startup as "global". But he saw an opportunity to build first in Shanghai because of the affordability of hiring engineers, as well as the size of the Chinese market and the resulting volume of unstructured data.

"The amount of unstructured data in a region is proportional to its population and level of economic activity, so it's easy to see why China is the largest source of data," he said at the time.

Of course, these days, with many startups in the country looking to relocate to other countries to have more freedom to grow their business and work with a wider range of clients, Zilliz is an example of this in action.

Prosperity7 has played a role in facilitating this migration. The fund entered China just last year and is actively seeking startups with global ambitions to capitalize on the company's extensive global network. We recently covered two such investments - Jaka, a Beijing-Shanghai startup in collaborative robotics, and Insilico, an artificial intelligence platform for medicines from Hong Kong.

Prosperity7's investment in Zilliz is well in line with the investor's objectives. These days, Chinese SaaS companies entering the international market is not uncommon. Many are founded by Chinese entrepreneurs with international experience who may have spent several years testing the market at home and attracting venture capital funds that are increasingly interested in B2B projects as the B2C space becomes saturated. However, many of them face monetization challenges in China, where SME owners are still reluctant to pay for software subscriptions compared to their Western corporate counterparts.

"Zilliz's leadership in Milvus makes it a global leader in finding vector similarities in massive amounts of unstructured data," said Aysar Tayeb, CEO of Prosperity7 Ventures, in a statement. "We believe the company is well positioned to build a cloud-based platform on Milvus that will provide its customers with new and powerful business insights and results, similar to what data analytics platforms like Databricks and Snowflake have done for structured data. There is already 4 times more unstructured data than structured data, and that gap will widen as AI, robotics, IoT and other technologies merge the digital and physical realms."