Summary of Key Points
Recently, the AI industry has seen the emergence of another trillion-dollar unicorn: Databricks, a company specializing in data infrastructure. It has secured a strategic financing of $5 billion, driving its valuation up to $19 billion (approximately 1.3 trillion yuan), placing it among the top five most valuable unlisted technology companies in the world (right behind OpenAI, Anthropic, ByteDance, etc.). Its business may seem unappealing at first glance—helping companies manage and integrate data—but it is essential in the AI era. For companies to implement AI effectively, they must first organize their internal data, and Databricks serves as the “infrastructure provider” that solves this problem. The financing was initially planned to be only $1 billion, but investors were so enthusiastic that they offered a total of $15 billion. Databricks ultimately accepted $5 billion and even turned down an additional $10 billion in offers, demonstrating the high demand for its services.
1. The Financing Boom: $5 Billion Wasn’t Enough; They Turned Down $10 Billion in Offers
This financing round can be described as a “capital feast”:
- Exceeding Expectations in Scale: The planned amount was $1 billion, but investors called the CEO non-stop, with a total offer of $15 billion. In the end, Databricks raised $5 billion and rejected an additional $10 billion (about 67.3 billion yuan).
- Valuation Surge: The current valuation of $19 billion represents a 42% increase from $134 billion at the end of 2025.
- Starred Investors: Existing investor Coatue led the round, with participation from over 20 institutions including Blackstone, Abu Dhabi’s AI platform MGX, and T. Rowe Price. Sixth Street Growth, founded by a former Goldman Sachs chief investment officer, also joined the investment.
- A Remarkable Financing History: Databricks has raised funds in 13 rounds over its 13-year history, with four rounds in 2025 alone (totaling $20 billion). The $1 billion raised in January’s Series J round set a record for private tech companies, which also saw the rejection of an offer of $19 billion at that time. Netizens joked, “They’ve raised so much money that they’re almost running out of letters of the alphabet!” (The financing rounds range from Series A to Series J).
2. Unappealing but Essential: The “Data Butler” and “Infrastructure Provider” in the AI Era
Databricks’ business is practical and crucial for the success of AI:
- Origins: Founded in 2013 by a team from the Berkeley Spark project, it started as a cloud-based data processing tool for simplifying large-scale data calculations.
- **Evolution to a “Data Lake Warehouse”: It integrates various types of data (sales, customer information, production data, etc.) into a single platform that allows for both storage and analysis, eliminating the need to transfer data between multiple systems, thus saving costs and improving efficiency.
- AI Integration: In 2023, Databricks acquired MosaicML, a generative AI company, and later bought Okera (for data governance) and Arcion (for data integration). It now offers a one-stop solution for companies to manage and use data for AI training.
- Powerful Customer Base: It serves 20,000 organizations worldwide, including Adidas, Bayer, and MasterCard. More than 1,000 customers spend over $1 million per year, and over 100 customers spend over $10 million annually.
3. Why Is It So Valuable? The Essential Path to AI Implementation
In the AI era, data is the key resource, but companies often struggle to organize it effectively:
- Large Models Need Business-Centric Data: Generalized AI models like OpenAI may not understand a company’s specific customer preferences, but models trained with company-specific data can.
- **Databricks Addresses the “Last Mile”: It converts scattered company data into a format that AI can use directly, making it essential for successful AI implementation.
- A Critical Role: Just as water sellers were essential during the gold rush, data infrastructure providers are indispensable in the AI era. Investors see Databricks as a necessary component for companies to leverage AI.
4. The Comeback Against the Trend: From a “Follower” to Outpacing Snowflake
Databricks and Snowflake were once competitors in the same market, but the gap between them has widened:
- Past: In 2020, Snowflake went public with a valuation of $33.5 billion, while Databricks’ valuation was only $6.2 billion (one-fifth of Snowflake’s).
- Current Situation: Databricks generates annual revenue of $7 billion (an 80% increase), compared to Snowflake’s $4.47 billion (a 29% increase). Databricks’ valuation is now $19 billion, while Snowflake’s market value is $113 billion.
- Reasons for the Difference: Snowflake remains focused on traditional data storage and analysis, while Databricks has expanded into AI infrastructure, leading to faster growth and a higher valuation.
5. The Future: The “Wealth-Creating Story” of AI Infrastructure
Databricks’ success shows that in the AI era, not only large AI companies (like OpenAI) but also those providing essential infrastructure are valuable. Its growth path (from research projects to data services to AI infrastructure) suggests to entrepreneurs that by addressing key challenges (such as data integration), they can reap the benefits of the current trend. The next Databricks could emerge from a company solving a specific infrastructure issue in the AI space.
In summary, Databricks’s story demonstrates that in the AI industry, companies providing essential infrastructure (such as data management and integration) can become trillion-dollar giants, just as much as those developing large AI models. This highlights the importance of underlying infrastructure in the value chain of AI, which extends beyond the front-end models.