① Databricks has launched Genie One, an AI agent for enterprise users that enables employees to obtain answers from internal data and support decision-making; ② Databricks also simultaneously introduced Genie Agents and Genie App Builder, along with Genie Code, a coding-focused AI agent tailored for developers.
Reuters, June 16 (Editor: Niu Zhanlin) — On Tuesday local time, data analytics and artificial intelligence software company Databricks announced the launch of an AI agent for enterprise users, enabling employees to derive answers from internal data and assist in decision-making. This move marks the company’s accelerated shift from traditional data services toward enterprise-grade AI.
The San Francisco-based company, valued at $134 billion, has named its new product Genie One, also referred to as an “intelligent collaborative colleague,” which helps business teams—including finance, marketing, and sales—derive insights and make decisions based on enterprise data.
Databricks’ product launch comes as data infrastructure providers increasingly position themselves as enterprise AI service providers. Snowflake, a key competitor of Databricks, exemplifies this trend. At the end of May, Snowflake’s stock surged following the release of strong earnings results and disclosures of growing demand for its AI offerings.
Like Snowflake, Databricks primarily provides cloud-based data storage and management infrastructure for enterprises. Since the rise of generative AI, such data platform companies have consistently emphasized that their proprietary data offers critical contextual information for AI models, thereby enabling more effective enterprise AI adoption—a proposition that is gaining increasing recognition among businesses.
Ali Ghodsi, co-founder and CEO of Databricks, stated that the company’s AI product business now generates over $1.7 billion in annualized revenue, a significant increase from the over $1 billion reported in September last year.
Ghodsi asserted that the core competitive advantage of the next-generation AI agents stems from a data context layer called “Genie Ontology.” This system functions essentially as a knowledge graph encompassing all of an enterprise’s knowledge assets, capable of integrating enterprise data, content, applications, documents, and employee information in real time.
He noted that a rich and continuously updated contextual environment enables AI to deliver more accurate and faster responses while reducing token costs during model inference.
In addition to Genie One, Databricks also unveiled two new products on Tuesday—Genie Agents and Genie App Builder—primarily targeting business users and enabling them to independently build AI agents and applications through “ambient programming.” Simultaneously, Databricks launched Genie Code, a coding-focused AI agent designed specifically for developers.
With the launch of these products, Databricks has officially entered the highly competitive AI agent market, where numerous AI agents targeting software development and knowledge work are already available.
In response, Ghodsi stated that Databricks’ agents are not intended to replace products such as Anthropic’s Claude Code, but rather to complement them. However, he anticipates that the AI agent market will become increasingly specialized in the future. "I believe specialization is inevitable. Data is our core strength, so we are focusing on this area," he said.
U.S. supermarket chain operator Albertsons and electric vehicle manufacturer Rivian have become early adopters of Databricks’ new product. Karthik Iyer, Head of Merchandise Operations Transformation and Artificial Intelligence at Albertsons, said the company is leveraging the technology to answer complex business questions based on internal enterprise data.
Romit Jadhwani, Senior Director of Enterprise AI, Data, and Productivity at Rivian, noted that Databricks’ agents enable employees to extract value directly from enterprise data without writing database query code. He revealed that Rivian’s management team is already using these agents to analyze key business data, including demand forecasts, production operations, and financial performance metrics.
As one of the most closely watched potential IPO candidates in recent years, Databricks’ listing plans have drawn significant investor attention. In this regard, Ghodsi said the company is closely monitoring major IPOs emerging in the capital markets this year, but Databricks is unlikely to go public in 2024.
According to media reports, Databricks is discussing a new round of fundraising, which could commence as early as next month. Sources familiar with the matter indicated that the company’s valuation in this round could reach between $165 billion and $175 billion, higher than its current valuation of approximately $134 billion.
Editor/Liam