We Chose Meilisearch Over 10+ Other Search Engines Despite a Major Drawback

Due to product growth and complex search needs, Hexmos is considering implementing a third-party search engine for their Feedback product. There are numerous search engines available, each catering to specific requirements and infrastructure capabilities. For small datasets and limited resources, PostgreSQL Full-Text Search, Bleve, and Tantivy are suitable options. For larger datasets and robust performance, server-based search engines like Meilisearch, Typesense, PISA Search, and Manticore offer scalability and advanced search operations. After evaluating several options, Hexmos selected Meilisearch due to its search-as-you-type and facet search capabilities, despite its higher RAM usage. Meilisearch also supports multiple schemas and typo handling, while Typesense has low CPU usage and high concurrency handling. PISA Search and Manticore lack search-as-you-type and facet search features, making them unsuitable for Hexmos's requirements. The decision process involved analyzing requirements, infrastructure capabilities, and the need for complex search operations. Hexmos plans to integrate Meilisearch to enhance navigation, search for user profiles, teams, and performance, and support future search needs across its product chain.