Where can I read more about the Prisma data proxy? Is it basically a hosted version of the query engine so that in serverless environments each function isn't spawning it?

Hello @thdxr, documentation for the proxy can be found here. Please note that this is still WIP as the proxy is still under development and in early access. The current implementation of the proxy involves a number of things: • the Query Engine is hosted indeed, so the connection pool is centralized • Prisma Client will connect to the Proxy via HTTP and the persistent connections are handled by the connection pool in the proxy • Since the Query Engine is not executed locally, the Rust binaries are no longer bundled, leading to smaller bundle sizes The improvements one can expect from it today is not exhausting the database connections with concurrent functions each opening connections, improved latency because of connections being reused and smaller bundle sizes. But we are also planning to evolve the product down the line and add more capabilities. Does this help? Please let me know if you have any follow-up questions. Also, we are still looking for users for our Early Access Program, so if you want to join, please sign up here.

Perfect exactly what I was hoping it was