No. Pulsbase automatically generates and manages vector embeddings for you in the background when you insert or update data. You just write SQL, and we handle the semantic understanding.

Absolutely. Pulsbase is fully SQL compatible. You can write traditional JOINs, GROUP BYs, and aggregations, and seamlessly mix them with our `SIMILARITY()` semantic search functions.

It allows you to query your database using plain English. Pulsbase safely translates your intent into an optimized, parameterized SQL query without risking SQL injection or exposing your raw data to LLMs.

Pulsbase is built on Cloudflare's global edge network. Your data is automatically replicated and distributed to over 330 cities worldwide, ensuring sub-50ms latency regardless of where your users are.