Citi: Databricks Conference Confirms Accelerating Enterprise AI Diffusion, Strongly Recommends Four Stocks: Microsoft Corporation (MSFT.US), Snowflake (SNOW.US), Palantir (PLTR.US), MongoDB (MDB.US)
Citi noted that enterprise AI diffusion is accelerating, while cost optimization has become a concern; demand read-through for Microsoft, Snowflake, Palantir, and MongoDB is positive, and it maintains Buy ratings on all four stocks.
Citi summarized key takeaways from the Databricks Data + AI World Tour event in New York, with its core judgment being that enterprise AI diffusion remains early but is accelerating; at the same time, cost optimization has become a persistent undercurrent throughout.
Citi maintains a "Buy" rating on all four companies in its coverage: Microsoft Corporation (MSFT.US) with a price target of US$680, MongoDB (MDB.US) with a price target of US$565, Palantir (PLTR.US) with a price target of US$265, and Snowflake (SNOW.US) with a price target of US$490.
Citi noted that customer and partner conversations around Databricks indicate that data and AI consumption is accelerating, driven by Genie, multi-agent architectures, and ongoing legacy platform migration. Genie is a key growth engine, with customer and partner feedback showing its advantages are particularly prominent in buyer scenarios such as the CFO's office; Unity Gateway, meanwhile, has won favor among enterprises managing multiple models and high token usage by integrating model access, routing, monitoring, and spend controls.
A set of data outlines progress on the adoption side: more than 70% of data pipelines have already been written by Genie Code, Genie's accuracy on real workloads has exceeded 80%, and active users within some organizations have reached hundreds of thousands. On the performance front, the Reyden engine has reduced query latency by up to 50% for a financial services customer during private preview, with general availability expected in January next year. After a large media company adopted a machine learning feature store supported by Lakebase, the same team tripled the number of models trained and increased experiments sevenfold.
But Citi also flagged that the cost optimization theme is ubiquitous. A FinOps partner managing approximately US$400 million in Snowflake and Databricks spend revealed that customers can achieve automatic savings of about 20% on Databricks, higher than the 10% to 13% on Snowflake, mainly from shutting down or right-sizing inefficient resources; traditional data workloads remain the largest savings pool. Another migration partner described how a large media customer's spending mix has shifted from 90% Snowflake and 10% Databricks a few years ago to a 50-50 split today, with Snowflake spending relatively flat. Citi believes the cost review and optimization trend warrants caution, especially concentrated among AI-native vendors and large language models.
On the competitive landscape, Citi noted the coopetition between Databricks and Palantirsome customers use Databricks to manage core data and Palantir to host ontologies or applications, and Palantir's explicit modeling ontology still holds a competitive advantage over Databricks' inferred ontology.
Overall, Citi believes that although cost optimization may temporarily weigh on consumption growth, the accelerating trend of enterprise AI diffusion and the implementation of multi-model, multi-agent architectures still bring positive demand signals to the software sector.
Related Articles

CHINA JINMAO (00817) subsidiary plans to publicly list for sale 100% equity of Jinmao Shenzhen Hotel at a reserve price of approximately RMB1.563 billion

SCHOLAR EDU (01769): Trustee purchases 155,000 shares under share award scheme

Changchun High-Tech Industry (000661.SZ)'s subsidiary's clinical trial application for GenSci161 Injection has been accepted.
CHINA JINMAO (00817) subsidiary plans to publicly list for sale 100% equity of Jinmao Shenzhen Hotel at a reserve price of approximately RMB1.563 billion

SCHOLAR EDU (01769): Trustee purchases 155,000 shares under share award scheme

Changchun High-Tech Industry (000661.SZ)'s subsidiary's clinical trial application for GenSci161 Injection has been accepted.






