OceanBase launches AI database portfolio with LakeBase

OceanBase has launched an AI database portfolio centred on a unified LakeBase architecture.

The portfolio combines three products: LakeBase as the underlying data engine, DataStudio for data production, governance and services, and DataPilot, a natural-language business intelligence tool. The offering is intended to help companies manage structured, unstructured and vector data in one system.

The move reflects a broader push by database suppliers to adapt their products for AI agents that need access to live business data rather than static records. OceanBase argued that many companies still rely on fragmented systems, making it harder for AI software to retrieve consistent information across documents, logs, images and transactional datasets.

Under the LakeBase model, OceanBase combines data lake storage, database functions and native multimodal processing in a single architecture. The aim is to give AI applications access to real-time context while preserving strong consistency across the underlying data foundation.

OceanBase has organised the portfolio into separate layers. LakeBase handles management, processing, search and serving for different data types, while DataStudio covers ingestion, orchestration, semantic modelling and agent collaboration. DataPilot is designed to generate reports, dashboards and answers from natural-language prompts for non-technical users.

OceanBase said enterprise AI deployment has slowed after rapid advances in large language models because businesses still struggle to give AI systems trusted access to operational data. That challenge becomes more acute as AI agents move from chat-based tools into workflow systems that require memory, state and continuous access to changing information.

A central claim in the launch is that companies can reduce complexity by replacing multi-system environments with a single foundation. OceanBase said its AI database can cut total cost of ownership by about 30% to 50% compared with traditional architectures.

It linked that claim to deployments in AI application scenarios at Ant Group, including AQ and Lingguang. According to OceanBase, Lingguang users have generated tens of millions of "flash apps", which it cited as evidence that the system can support isolated data environments at large scale.

Industry use

OceanBase said its AI database products have already been adopted in several sectors. It named Lalamove, China Unicom and Trip.com among companies using its software for retrieval-augmented generation, hybrid search and AI-based data applications.

Those use cases point to a commercial opportunity for database groups seeking a role in enterprise AI stacks. Businesses are increasingly looking for systems that can combine analytics, search and transactional records without forcing teams to move data repeatedly between separate platforms.

OceanBase positioned consistency, scalability, reliability and real-time performance as the core requirements that remain unchanged even as data management shifts towards AI-led workloads. The company said its approach extends features already used in financial core systems to broader multimodal and data lake environments.

Charlie Yang, Chief Technology Officer at OceanBase, outlined that position in the launch announcement. "As AI moves from answering questions to taking actions, databases must evolve from systems of record into trusted context engines for AI," Yang said.

He added: "OceanBase AI Database is not about stitching together data lake and database. It is about bringing multimodal data, real-time serving, transaction consistency, and open compute into a single architecture."