🔗 Introducing TIN: Full-text search for PostgreSQL

One of the most sought-after features among Postgres users is full-text search, leading to the launch of TIN—"Text INdex." This new GA release promises a fast and comprehensive full-text search experience for all Postgres and Neki databases. TIN stands out by supporting Boolean expressions, phrase queries, case and accent folding, and BM25-scored queries. Importantly, TIN outperforms existing text-search indexes in terms of speed and reliability, managing to handle a high volume of queries concurrently, even amidst document updates.

Eric Ridge and Patrick Reynolds detail how rigorous benchmarking was conducted, demonstrating TIN's remarkable capabilities across various workloads, including a mix of academic papers and user-generated content. Results showcase TIN handling up to 25x more queries per second when compared to alternatives, with substantially reduced latency. This is a promising tool for developers looking to enhance search functionalities in their applications, from e-commerce to legal discovery platforms.