Overview

This section is for the teams running K-AI in production — platform / SRE teams, the Document Engineer (DocOps) practice. It covers deployment, observability, scaling, and the most common operational situations.

For architectural context, see The K-AI Platform. For the commercial overview of deployment modes, see Platform — Deployment models.

What's in this section

• Deployment models — ops-grade view of SaaS, on-premise, and Snowflake Native App: substrate, isolation model, responsibilities.

• Monitoring & observability — what K-AI emits, how to consume it, what you can observe self-service.

• Scaling & quotas — how K-AI scales, what's physically isolated, which quotas are configurable.

• Common issues — self-service guidance for the five operational situations customers see most often.

• On-premise installation — prerequisites, air-gapped flow, install model, customer / K-AI responsibility split.