Skip to content

Model Engine — Operators Guide

Internal documentation for service owners and deployment engineers installing, operating, and debugging model engine in customer environments.

Not the end-user docs? The user-facing Python client documentation is at the root of this site.


Contents

Document What it covers When to use it
Architecture System overview, component deep-dives, lifecycle flows, autoscaling Before your first deployment; when debugging unfamiliar failures
Helm Values Reference Every configurable value, organized by concern, with high-risk callouts During installation and upgrades
Smoke Tests Post-deploy validation checklist (Tier A: CPU, Tier B: GPU+LLM) After every helm install or helm upgrade
Cloud Support Matrix Per-cloud config reference, behavior differences, image mirroring When deploying to a specific cloud


Contributing

These docs live at docs/internal/ in the repo. Update them when you change behavior or configuration — the PR template includes a reminder.

Rule: If a PR author would need to update this doc when changing the code → it belongs here. Operational notes from specific deployments belong in Confluence.