Data Infrastructure Becomes the Real AI Battleground
Enterprise data infrastructure for AI refers to the databases, data platforms, and governance tools that prepare, store, and deliver corporate data so machine-learning models, generative systems, and AI agents can work with it reliably at scale. The soaring valuations of ClickHouse and Databricks show this layer is turning into a primary destination for enterprise AI spending. Instead of focusing only on model access, companies are investing in AI data platforms that can handle both structured and unstructured information, feed agents with low-latency analytics, and enforce governance across business domains. That shift reflects a simple lesson from early deployments: without clean, accessible, and timely data, even the most advanced models fail to deliver business value. As budgets grow, boards and CIOs are treating data infrastructure valuations as a signal of which platforms are becoming core to long-term AI strategy.
ClickHouse’s Revenue Surge Signals IPO-Ready Data Infrastructure
ClickHouse’s traction shows how fast modern databases tied to AI agents are maturing into large businesses. The company has surpassed USD 250 million (approx. RM1,150,000,000) in annualised revenue run rate, tripling year-over-year, and its president Yury Izrailevsky expects that figure to reach the high nine digits by year end. That growth supports a USD 15 billion (approx. RM69,000,000,000) valuation from a recent Series D and has pushed ClickHouse to prepare for an IPO, including hiring former Snowflake investor relations chief Jimmy Sexton as CFO. With more than 4,000 customers such as Anthropic, Meta, and Capital One, its open-source database is tuned for the massive datasets that AI agents consume. Managed cloud services drive most of its commercial revenue, while acquisitions like Langfuse extend into monitoring and evaluating AI agent performance, reinforcing ClickHouse as a central piece of enterprise data infrastructure.
Databricks’ $134B Price Tag and the AI Data Platform Wars
Databricks has become a bellwether for data infrastructure valuations in AI. According to an Inc. profile cited by ERP Today, the company recently closed more than USD 7 billion (approx. RM32,200,000,000) in new funding, including USD 5 billion (approx. RM23,000,000,000) in equity at a USD 134 billion (approx. RM616,400,000,000) valuation. Earlier this year it reported an annual revenue run rate of USD 5.4 billion (approx. RM24,840,000,000), with fourth-quarter revenue up more than 65% year over year, an acceleration that investor Tomasz Tunguz called “exceptional” at this scale. Databricks’ Data Lakehouse model unifies structured and unstructured data, positioning the company in a head-to-head contest with Snowflake to define how enterprises prepare, govern, and operationalize data for AI. The fight is no longer about storage alone; it is about owning the AI data platform layer closest to business execution and analytics.
Enterprise AI Spending Shifts from Models to Data Platforms
Early enterprise AI investments concentrated on accessing frontier models, but spending patterns now show a pivot toward AI data platforms. Databricks and Snowflake are competing to become the default environment where corporate data becomes usable for analytics, automation, and agentic AI, while ClickHouse anchors high-performance querying for massive datasets. For ERP and application leaders, this means that data architecture is now a central part of AI strategy: finance, supply chain, HR, and customer operations all depend on combining structured ERP data with documents, messages, and media. As organizations confront AI bottlenecks, the biggest constraints are data quality, accessibility, and governance rather than model choice. Budgets are following that reality, directing enterprise AI spending into platforms that can prepare, secure, and deliver data to many models, not locking into a single model vendor. This rebalancing is redefining which technology categories sit at the center of AI roadmaps.
IPO Paths and Mega-Rounds Cement Data Infrastructure as Core Stack
The capital flows around ClickHouse and Databricks underline how enterprise data infrastructure has become a core technology category, not a supporting tool. ClickHouse’s IPO ambitions, paired with its aggressive acquisition of open-source startups like Langfuse, show that even specialized databases can now reach scale and public-market readiness on AI demand. Databricks’ and Snowflake’s acquisition races, including deals for MosaicML, Arcion, Neon, Mooncake Labs, Tabular, and Neeva, highlight a broader land grab across model customization, governance, and real-time transaction data. These moves signal that the “AI stack” is solidifying around a powerful data layer that application vendors, ERP systems, and integrators increasingly depend on. As more enterprises standardize on a small set of AI data platforms, valuations and IPOs in this space are best seen not as hype, but as markers that data infrastructure now sits at the heart of digital and AI transformation programs.