en.Wedoany.com Reported - PhoenixAI Inc. (formerly CelerData) announced it has secured $80 million in new funding to advance the development of its AI-native database and expand governance for regulated industries. The Series B round was led by Sky9 Capital, with participation from Atypical Ventures, Olive Technology Ventures, and existing investors.
As AI agents move from planning and prototyping into production, their application patterns have shifted from experimentation to large-scale deployment. With massive numbers of AI agents submitting thousands of query requests per second to databases, the modern data stack is becoming overwhelmed. PhoenixAI has built an analytics database designed for agentic AI readiness, aiming to serve this era of agents and the growing needs of enterprises. Transactional databases handle individual operations, such as inserting rows, updating account balances, and recording orders. They reliably and atomically record one thing at a time, with persistence and highly normalized data. Agents struggle here because these data formats are optimized for narrow operations, while agents deal with unstructured, conversational use cases that mimic human thinking, such as "Who are our top 10 customers by revenue growth over the past 90 days, broken down by product line?" — an analytical question that requires scanning millions of rows across multiple tables.
Analytical databases can store and process complex queries on massive datasets. They scan billions of rows, performing aggregations, joins, and summaries, but sacrifice write speed for read speed. Representatives in this field include Snowflake Inc., ClickHouse Inc., Apache Druid, Google BigQuery, and PhoenixAI. Analytical databases do not replace transactional databases; both coexist. Transactional databases serve as systems of record, acting as the source of truth, while analytical databases provide systems of insight for the agent world. PhoenixAI is not looking to replace enterprise Oracle or SAP ERP systems but to build a layer on top of them to make them smarter, enabling AI agents to act and think faster.
PhoenixAI stated that it has re-architected its analytical database to meet the demands of the agentic AI era, particularly handling clusters of thousands of agents seeking information. President Rick Underwood noted that most existing analytical databases were designed for a world that no longer exists—a world where humans ran dashboards on flat tables and complexity was someone else's problem. When thousands of agents need to query, reason, and act simultaneously on petabytes of real-time data—any question, simple or complex—the database is either the bottleneck or the breakthrough. PhoenixAI claims to achieve sub-second latency and high concurrency on real-time data, enabling multiple agents to query simultaneously. This allows agents to ingest data quickly while it updates across large scales—no more waiting, blocking, or bottlenecks. The company calls this form "pipeline-free," where fresh data from Kafka (an open-source event streaming platform for decoupling data pipelines) flows in continuously, updating information within seconds rather than minutes or hours.
Other major players in the analytical database market are also advancing their capabilities. Snowflake has just launched its own agent features. Databricks Inc. is pushing toward real-time processing with Delta Live Tables. ClickHouse Cloud has significantly improved concurrency. A race is underway to consolidate the "agentic database" market category and build the infrastructure data layer that will fuel this voracious AI future. In this future, queries depend not only on which database row is checked but also on how analysis interacts with information.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
