Case Study: CommSync achieves 1,000x faster search with Neon Lakebase Search

A Neon Case Study

Preview of the CommSync Case Study

Commsync improves search latency 1,000x with Neon

Neon's customer Commsync, a platform providing a unified inbox for business communications, faced a significant performance challenge with its core search functionality. Their existing Postgres-based GIN index implementation was too slow, especially on cold starts, creating a poor user experience across their high-traffic inbox search, AI assistant, and new command palette feature.

The solution was implemented using Neon's Lakebase Search extensions, lakebase_text and lakebase_vector, which run directly on Postgres. This allowed Commsync to power all three search surfaces—text, vector, and hybrid—within their primary database without a separate search stack. The results were substantial, with Neon reporting a 1,000x improvement in query latency, reducing search times from approximately 19.5 seconds to just 18.6 milliseconds, while also simplifying their architecture and permission model.


View this case study…

Neon

39 Case Studies