Hybrid Search: OLAP with vector search and full-text search and SQL analytics (www.velodb.io)

Apache Doris has officially launched version 4.0, marking a significant advancement with a focus on integrating AI capabilities, enhanced search functionalities, improved ETL/ELT processes, and performance optimization. The release introduces vector indexing, allowing users to perform both vector search and standard SQL analytics within a single platform, negating the need for external vector databases. This hybrid capability enables complex queries, including precise keyword searches and semantic matching, which are crucial for applications like smart recommendations and image retrieval. Notably, Doris 4.0 enables data analysts to utilize various AI functions via SQL, allowing direct interaction with large language models for tasks such as information extraction and sentiment analysis. The new SEARCH() function simplifies full-text searching with a DSL syntax akin to Elasticsearch, enhancing flexibility and performance. Additionally, performance improvements, such as lazy materialization and SQL caching, promise up to a 100x boost in parsing efficiency and faster execution of wide-table queries, contributing to a more robust and capable tool for data-intensive applications. Overall, this release underscores Doris' commitment to providing a unified platform that caters to increasingly complex data and AI workloads.