Hi all, is anyone here an expert at optimizing custom models. For example with ONNX and/or TensorRT?

We did onyx a lot at a previous company but it's been awhile! I'm actually looking into using Onyx,TensorRT, or RTlite for some projects right now though!