In 2020, Chinese startups Ziliz — who builds cloud-native software to process data for AI applications and unstructured data analytics, and is the creator of milvusPopular open source vector database for common ground — raised $43 million to scale its business and prepare the company to move to the U.S. Nearly two years later, Zilliz today announced a further $60 million in funding as it finally moves to the West, with new headquarters in San Francisco to capitalize on the growing demand for more efficient processing techniques for the ever-expanding unstructured data set commanded to support AI applications.
Funding led by Prosperity7 Ventures, a $1 billion venture fund made by oil giant Saudi Aramco (the name refers to the country’s first oil-producing commercial well), with backers China Pavilion Capital, Hillhouse Capital, 5Y Capital and Yunqi Capital – all previous investors – also participating. The company did not disclose its valuation, but it is worth pointing out that this latest injection is described as an extension of that $43 million Series B rather than a new round. We will update this as we learn more. The total raised by the company is now $113 million.
The capitalization and relocation speaks not only to a pivotal moment for the company, but also to the broader area of machine learning and trends impacting start-ups established in China.
The first, Zilliz’s breakthrough product, the open-source Milvus, has grown rapidly. The company says that downloads have now crossed the 1 million mark, compared with 300,000 last year, with production users growing 300% in the same period, though it did not disclose the number of active users. Customers include the likes of eBay, Tencent, Walmart, Ikea, Intuit and Compass.
As we demonstrated in 2020, Milvus relies less heavily on advertising and marketing spending, instead opting to leverage word of mouth in places where developers like to congregate for ideas and inspiration to get themselves noticed, such as Github and Reddit. . The strategy worked: “Stargazers” on Github were up 200% to over 11,000 with the number of contributors doubling. (For comparison, in 2020 it has been starred about 4,400 times.)
The reason for the interest in Milvus — and hence the Zilliz roadmap, which is based on the creation of further products, the latest of which is Zilliz Cloud managed services that are now in private preview; and Towhee, another open source framework, this one is for ETL vector data — due to the growing interest in vector databases such as how they are used in AI applications.
Simply put, while data can (and often is) processed through more traditional databases, the complexity and structure of activities such as anomaly detection, recommendation, assessment, and other AI-based tasks lend themselves to vector databases designed to work more naturally and efficiently. with how AI data is represented. (Zilliz notes that its vector database “comes from the cloud and is capable of processing billions of scale vector data in milliseconds.”)
“Ziliz’s journey to this point began with the creation of Milvus, an open source vector database that eventually joined the LF AI & Data Foundation as a top-level project,” said Charles Xie, founder and CEO of Zilliz, 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 serve as key contributors and committers to Milvus and deliver on our promise to provide fully managed vector database services in the public cloud with the security, reliability, ease of use and affordability enterprises need.”
There are others (competitors) like pinecone and weave also build solutions to address this; and major cloud providers like AWS also has its own solution. All of these indicate market opportunities that Zilliz is focused on addressing.
It’s worth noting that in 2020, Zilliz has said that more than half of Milvus’ users are outside China: that shows how the company has long positioned itself and where it sees growth in the long term. Today, a number of startups in the country are looking to move elsewhere to have more freedom in how they grow their business, and to work with more customers, and Zilliz is an example.
Prosperity7 has been instrumental in facilitating this migration. The fund only entered China last year and has been actively pursuing startups with global ambitions, who can take advantage of the company’s vast global network. We recently covered two of those investments, Jaka, a collaborative robotics startup based in Beijing and Shanghai, and Insilico, an AI medicine platform from Hong Kong.
Prosperity7’s investment in Zilliz appears to live up to the investor’s mandate. It’s not uncommon to see Chinese SaaS companies going global these days. Many were started by Chinese entrepreneurs with international backgrounds. They may have spent several years testing the market at home and upgrading from VCs who are increasingly interested in B2B projects as the B2C space becomes saturated. But many find it difficult to make money in China, where small and medium-sized business owners are still reluctant to pay for software subscriptions compared to their Western corporate counterparts.
“With his leadership at Milvus, Zilliz is a global leader in vector similarity searches on large amounts of unstructured data,” said Aysar Tayeb, executive MD of Prosperity7 Ventures, in a statement. “We believe the company is in a strong position to build a cloud platform around Milvus that will bring out new and powerful business insights and results for its customers, just as data analytics platforms like Databricks and Snowflake have done with structured data. There is already more than 4x more unstructured data than structured data, a gap that will continue to grow as AI, robotics, IoT and other technologies blend the digital and physical realms.”