I want to know what exactly is the inference pipeline written on?

Since python is not synonymous with speed how roboflow aims to make or extract the speed for the inference pipeline.
is the inference pipeline a wrapper for the deepstream? ( I don’t think so)
For the development environment the inference pipeline produce the desired results and is relatively easy to set it up but when it comes to the production grade computer vision project deployment how does it compare? Any benchmarks that I am not aware of?

1 Like

no response :frowning:

Hi, it’s open source here; you’re welcome to have a look.

It’s made to be used in production and several of the world’s largest companies rely on it in their operations. Please let us know if you have specific questions/issues.