Webinar | Faster and More Powerful Full-Text Search with Milvus 3.0

Search applications often need both exact keyword matching and semantic understanding. Traditionally, delivering both meant operating separate full-text and vector search engines, maintaining duplicate write paths, reconciling different scoring models, and implementing complex result-fusion logic in the application layer. Milvus 3.0 closes this search gap by bringing full-text search and vector search together in one engine. With native BM25 scoring, hybrid retrieval, flexible schemas, kernel-side execution, and more efficient indexes, teams can build faster and more powerful search experiences without synchronizing multiple systems or combining results by hand. In this webinar, Simon Hearne, Solutions Architect at Zilliz EMEA, explains how Milvus has evolved into a unified engine for semantic and full-text search. He examines the limitations of traditional two-engine architectures, introduces the search and schema capabilities available in Milvus 3.0, and demonstrates Milvus and Elasticsearch side by side using the same Amazon Reviews dataset, vector embeddings, analyzer configuration, and compute resources. If you are a search engineer, AI application developer, data architect, or platform leader building RAG systems, AI agents, legal discovery tools, observability platforms, e-commerce search, or other retrieval-intensive applications, this session provides a practical look at how Milvus 3.0 can simplify your architecture while improving search performance, flexibility, and cost efficiency. What you'll learn: ✦ Why combining a traditional full-text search engine with a separate vector database creates operational and relevance challenges ✦ How native BM25 and vector search work together in Milvus to produce a single ranked result ✦ Where hybrid keyword and semantic retrieval adds value across RAG, AI agents, legal and compliance, observability, and e-commerce ✦ How teams are consolidating Elasticsearch or OpenSearch workloads onto Milvus and Zilliz Cloud ✦ What Milvus 3.0 adds across full-text search, hybrid retrieval, grouping, aggregations, relevance shaping, and search efficiency ✦ How custom dictionaries, multilingual analysis, highlighting, phrase matching, and geo filtering improve search experiences ✦ How nullable fields, JSON shredding, path indexes, and dynamic fields support sparse and evolving real-world data ✦ How kernel-side execution, efficient vector quantization, smaller BM25 indexes, and the Loon storage engine improve performance and resource efficiency ✦ How Milvus 3.0 and Elasticsearch 9.4 compare in a controlled test using identical documents, vectors, analyzers, and compute limits ✦ How keyword search, phrase matching, highlighting, hybrid retrieval, grouping, and schema flexibility work in a live side-by-side demo ✦ Which additional capabilities—including fuzziness, phrase-prefix search, multi-stage rescoring, and Bloom-filter support—are planned for the future ✦ How to reproduce the complete Milvus and Elasticsearch comparison locally with Docker Slides from this webinar are available here: https://24054828.fs1.hubspotusercontent-na1.net/hubfs/24054828/Webinars/webinar%20FTS%2008192026.pdf Run the demo yourself: https://github.com/simonhearne/milvus_es_lab Find us on: Zilliz Website: https://zilliz.com Milvus Website: https://milvus.io Zilliz LinkedIn: https://www.linkedin.com/company/zilliz/ Milvus LinkedIn: https://linkedin.com/company/67143129 Milvus X: https://x.com/milvusio Milvus GitHub: https://github.com/milvus-io/milvus Try Zilliz Cloud: https://cloud.zilliz.com/signup