Search Your Lance Table Faster with Milvus External Collection

About the Webinar

Your embeddings and metadata already live in Lance tables. Adding vector search often means copying that data into another system and maintaining a pipeline to keep both copies aligned.

In this webinar, we’ll explore how Milvus External Collection indexes and searches data in Lance tables while the source data stays in place. We’ll walk through connecting an external source, mapping its columns to Milvus fields, building indexes, and running vector searches with metadata filters. We’ll also explain how refresh makes source changes available for retrieval and when an external collection fits your workload.

What You’ll Learn

• Connect Lance Tables to Milvus — Understand the Storage V3 requirements and map source columns to a Milvus collection schema.
• Index and Search External Data — Build vector and scalar indexes, then search with metadata filters through the standard Milvus APIs.
• Manage Data Freshness — Learn how explicit refresh makes updated source data available for search.
• Understand the Tradeoffs — Explore read-only access, index and cache requirements, and when to choose an external collection or a regular Milvus collection.
• Live Q&A — Bring your questions about Lance tables, external collections, and vector search over lake-resident data.