Zilliz has announced the public preview of Zilliz Vector Lakebase, a Zilliz Cloud release that combines vector search, shared lake-native storage, on-demand compute, external data lake search and batch analytics for AI applications.
Zilliz announced the public preview of Zilliz Vector Lakebase, a Zilliz Cloud release intended to combine vector search with lake-native storage and analytics capabilities for AI data infrastructure.
According to PR Newswire’s release from Zilliz, Vector Lakebase brings together production vector search, shared lake-native storage, on-demand compute, external data lake search and batch analytics. AI-Watch, republishing the PRNewswire item, described the release as a public preview that combines real-time serving, discovery and batch analytics for AI data infrastructure.
In a Zilliz blog post titled “From Vector Database to Vector Lakebase,” the company said the product is meant to extend Zilliz Cloud beyond a vector database into a broader platform for managing AI data. Zilliz describes the service as using an S3-based unified data foundation to support real-time retrieval, iterative discovery and batch analytics.
The announcement positions Vector Lakebase as a response to a common infrastructure problem in AI systems: vector search, offline analysis and data lake storage often live in separate systems. Zilliz says its approach is to let teams work from shared lake-native storage while using compute as needed for serving and analytics.
Vector databases are widely used in retrieval-augmented generation, semantic search, recommendation systems and other applications that compare embeddings rather than relying only on keyword matching. Zilliz is the company behind Milvus and Zilliz Cloud, and the PR Newswire announcement describes its vector database technology as widely adopted. That adoption claim comes from Zilliz’s own announcement and should be read as company positioning rather than an independently verified market ranking.
The notable part of the Vector Lakebase preview is the attempt to merge several stages of AI data work. In the PR Newswire release, Zilliz says the release supports production vector search alongside shared storage and batch analytics. In its blog post, the company frames this as a move from a standalone vector database model toward a lakebase model, where real-time retrieval and longer-running analysis can operate on a common data foundation.
That could be useful for teams building AI applications that need both low-latency retrieval and offline evaluation or exploration. For example, an application may need to serve relevant documents to a model in real time, then analyze retrieval quality, embedding drift or dataset coverage in bulk. Zilliz’s announcement suggests Vector Lakebase is designed to reduce the need to move data between separate tools for those tasks.
The release is currently a public preview, according to Zilliz and PR Newswire. That means enterprises evaluating the product should treat it as an early-access offering rather than a fully mature general-availability service. The source materials do not provide pricing, service-level commitments or detailed performance benchmarks for the preview.
Zilliz’s own blog emphasizes an S3-based unified data foundation, real-time retrieval, iterative discovery and batch analytics. The PR Newswire release adds on-demand compute and external data lake search to the list of supported capabilities. Together, those sources show the company is trying to broaden Zilliz Cloud from vector serving into a wider AI data platform.
The announcement fits a broader trend in AI infrastructure: vendors are trying to reduce the separation between storage, search and analytics as organizations put more embedding-based systems into production. Zilliz’s Vector Lakebase preview is one example of that shift, with the company betting that AI teams will want vector search and lake-style data operations in a single managed environment.
Zilliz announced the public preview of Zilliz Vector Lakebase, a Zilliz Cloud release intended to combine vector search with lake native storage and analytics capabilities for AI data infrastructure.
AI Watch, republishing the PRNewswire item, described the release as a public preview that combines real time serving, discovery and batch analytics for AI data infrastructure.
In a Zilliz blog post titled “From Vector Database to Vector Lakebase,” the company said the product is meant to extend Zilliz Cloud beyond a vector database into a broader platform for managing AI data.
Continue reading
