Skip to main content
Most AI material is written for data scientists and Python notebooks. It assumes you want to train models, tune hyperparameters, and plot loss curves. You almost certainly do not. You want to consume models through an API and wrap them in robust, observable, secure software that runs in a container next to everything else you operate. That is an engineering problem — your problem — and it has been badly underserved. This Docs closes that gap. It assumes you are a strong operator and a competent (not expert) Go Programmer, and it assumes no prior AI experience whatsoever. Every term is defined before it is used. Every acronym is expanded on first appearance. We never wave our hands and say “the model just figures it out”; we open the box and show you the gears, then show you how to build on top of them.
Last modified on June 8, 2026