Enterprise Data Platforms as the New AI Control Point
Enterprise data platforms are integrated systems that store, govern, and process structured and unstructured data so organizations can turn raw information into analytics, automation, and AI-powered applications. They now sit at the center of AI infrastructure competition because they control the layer where business data is cleaned, combined, and exposed to models. Instead of focusing only on model training, enterprises are shifting spending toward platforms that make AI model deployment reliable, observable, and economically sound. This is driving a consolidation of value around the platforms that bridge operational systems, analytics, and agentic AI. Databricks, Snowflake, and ClickHouse are among the most prominent examples, each racing to become the essential infrastructure layer that connects data pipelines, governance, and production AI workflows across finance, operations, and customer-facing systems.
Databricks’ Soaring Valuation and the Economics of AI Readiness
Databricks’ latest funding round, more than USD 7 billion (approx. RM32.2 billion) including USD 5 billion (approx. RM23 billion) in equity, pushed its data platform valuation to USD 134 billion (approx. RM616.4 billion). That came only months after a USD 1 billion (approx. RM4.6 billion) raise above USD 100 billion (approx. RM460 billion), signaling intense investor conviction that the data layer is where enterprise AI value concentrates. The company’s roots in Apache Spark and unstructured data gave it a head start once enterprises sought to plug internal documents, logs, and messages into AI systems. Databricks’ Data Lakehouse strategy, spanning structured and unstructured data, and its rapid revenue run-rate growth to USD 5.4 billion (approx. RM24.8 billion) underline that AI economics now favor platforms that make data usable, governed, and production-ready rather than tools focused only on building models in isolation.
ClickHouse’s Growth and Product-Market Fit in AI Analytics
ClickHouse is showing a different but complementary side of the enterprise data platforms story by anchoring itself in high-performance analytics and AI-era workloads. The company’s serverless cloud business has crossed USD 250 million (approx. RM1.15 billion) in annual run-rate revenue, more than triple year over year, with over 4,000 customers using it for high-concurrency queries and telemetry from AI applications. That expansion, accelerated by its USD 400 million (approx. RM1.84 billion) Series D, shows strong product-market fit for cost-efficient analytics tightly connected to AI systems. Customers range from financial services and e-commerce to autonomous systems and model observability tools, reflecting how widely ClickHouse now sits in the stack. Its growth reinforces that AI infrastructure competition is not only about general-purpose data warehouses, but also about specialized engines that can handle the intense, low-latency analytical demands of live AI model deployment.
AI Agents and the Shift from Data Warehousing to AI Systems
Both Databricks and ClickHouse are moving beyond traditional warehousing toward platforms tuned for agentic AI. ClickHouse Agents, powered by Anthropic’s Claude, is a fully managed, no-code agent builder that ties agents directly to ClickHouse data and MCP-compatible systems, with features like a sandboxed code interpreter, skills management, memory, and multi-agent workflows. Databricks, for its part, is positioning its Data Lakehouse as the central plane where analytics, automation, and AI agents share a single governed data foundation. These moves show a clear shift in enterprise spending priorities: from classic reporting to continuous AI model deployment, observability, and feedback loops. Enterprises want platforms that can support telemetry, agent-driven analytics, and AI observability on the same data, so that models can be monitored, evaluated, and tuned without stitching together fragile, expensive point solutions.
Consolidation of the Critical Layer Between Data and AI
As enterprise AI adoption matures, the competitive landscape is consolidating around platforms that sit between raw data sources and production AI systems. Databricks and Snowflake are vying to define how enterprises prepare and govern data for analytics and AI, while ClickHouse focuses on cost-effective, high-throughput querying and AI observability workloads. This layer now includes managed transactional stores, full-text search, and observability stacks built for model-training telemetry and agent correctness. The strategic prize is control over the environment where AI models are fed, monitored, and iterated, rather than the models themselves. For enterprises, this means AI infrastructure decisions increasingly revolve around which data platform offers the best combination of performance, governance, and AI-native capabilities. For investors, the rising data platform valuation levels show that this middle layer is likely to capture a large share of long-term AI spending.
