The rivalry between Snowflake and Databricks has expanded beyond data lakes and warehouses, moving into the growing world of PostgreSQL.
The battle has intensified with strategic acquisitions on both sides. Snowflake recently acquired Crunchy Data for $250 million, while Databricks snapped up Neon for $1 billion.
Announcing the acquisition in a blog post, Snowflake called Postgres a top choice for developers due to its flexibility, cost efficiency, and native AI features, such as vector support (pg vector). Its open-source nature and vibrant ecosystem only add to its appeal.
“We’re tackling a massive $350 billion market opportunity and a real need for our customers to bring Postgres to the Snowflake AI Data Cloud,” said Vivek Raghunathan, Snowflake’s SVP of engineering.
Why are the Giants Betting on PostgreSQL?
“PostgreSQL’s ecosystem and extensions are growing fast. More people now know this database better than any other. pgvector gave it a big push,” said Arpit Bhayani, creator of DiceDB, when AIM asked why PostgreSQL is becoming the preferred database for AI-native applications.
Snowflake Postgres builds on Snowflake’s earlier move into transactional data with Unistore, which combines transactional and analytical workloads in one system.
Building on native PostgreSQL support, Snowflake Postgres extends that vision, offering enterprises a production-ready solution for transactional applications that require Postgres compatibility.
PostgreSQL, the open-source relational database, has surged in popularity, surpassing MySQL as the most favoured database among developers, according to Stack Overflow’s 2023 and 2024 Developer Surveys.
Its ability to handle geospatial data (via PostGIS), time series data (via TimescaleDB), JSON, and vector embeddings (via pgvector) makes it ideal for AI applications.
Commenting on the acquisitions in a LinkedIn post, senior data engineer Avinash S said it’s more than just adding another database, calling it a strategic bet on PostgreSQL as the backbone of the AI-native era, especially in its serverless and cloud-native form.
“Imagine AI agents spinning up databases for every real-time task or experiment, then discarding them. Traditional databases can’t handle this “disposable” scale. Serverless Postgres delivers the rapid provisioning, elasticity, and cost-efficiency that AI agents desperately need to work autonomously and at speed,” he added.
“It’s not just agentic. Because many people are talking about it and using it, it has become the de facto standard,” said Bhayani. However, he added that AI agents spinning databases seems like a strong assumption.
Similarly, Factorial Advisors said in a blog post that the Neon acquisition fits into Databricks’ broader push to build a unified data intelligence platform.
“With over $19 billion in financing and a recent $62 billion valuation, Databricks has the financial muscle to keep acquiring,” they wrote. “Following previous buys like Tabular ($2 billion) and MosaicML ($1.3 billion), Neon helps address the growing demand for databases that operate at ‘agentic speed’ while staying cost-effective through pay-as-you-go economics.”
With these acquisitions, both Snowflake and Databricks are going to challenge hyperscalers like AWS, Microsoft Azure, and Google Cloud, which offer managed PostgreSQL services tightly integrated with their AI stacks.
Neon vs Crunchy Data
Founded in 2012, Crunchy Data provides a full-service, production-ready PostgreSQL solution that includes backups, high availability, disaster recovery, connection scaling, and monitoring. It supports mission-critical deployments across cloud, on-premises, and hybrid environments.
Snowflake said its new offering, Snowflake Postgres, will bring transactional Postgres data into its platform, accelerating innovation and giving developers greater agility, visibility, and control to build trustworthy AI agents and applications faster.
Crunchy’s expertise means Postgres-powered apps can run natively on Snowflake without having to rewrite code. Developers benefit from built-in connection pooling, performance metrics, and logging, making it easier to build and manage scalable apps.
On the other side, Databricks CEO Ali Ghodsi pointed out that frontier LLMs have been trained on vast amounts of public data from the Postgres open-source ecosystem, making AI agents naturally proficient in using Neon, which is built on Postgres.
Ghodsi added that Databricks and Neon share a common foundation in technical infrastructure and a deep belief in open source. Databricks started the Apache Spark project at UC Berkeley, also the birthplace of Postgres, the open-source database on which Neon is built.
He added that OLTP databases, a $100 billion market, are still dominated by decades-old products. With Neon, Databricks aims to help disrupt this space by creating the most developer- and AI agent–friendly database platform.
When Neon became generally available last year, around 30% of the databases on the platform were being created by AI agents rather than humans. But in a recent update, that number has jumped to over 80%. In other words, AI agents are now creating more than four times as many databases as human users.
Everyone’s rushing to PostgreSQL, and Snowflake and Databricks are jumping in by picking up niche providers. It’s not just about more databases, it’s about getting ready for AI, live data, and big business needs.
Moreover, these acquisitions reflect a broader consolidation trend in the data and AI infrastructure market.
Recent deals such as Salesforce’s $8 billion acquisition of Informatica, ServiceNow’s purchase of Data.World, and Alation’s acquisition of Numbers Station, show how companies are racing to build comprehensive AI-ready platforms. According to Bhayani, much of this is driven by the push to acquire customers and specialised expertise.
The post Snowflake and Databricks Take Their Rivalry to PostgreSQL appeared first on Analytics India Magazine.
