Skip to content

Guztia AI Infra Ops

Guztia /Services /AWS China

Cloud platform

AI Infrastructure on AWS China

A separate partition, separate accounts, separate ARNs, and an ICP filing before anything serves traffic. If you need inference that answers inside the mainland for mainland users, this is one of the two credible routes.

01 The short version

For companies that already run AWS globally and need a mainland presence that their existing tooling can mostly address. The partition boundary is real — plan for duplicated IaC, not a copied stack.

  • Operated by Sinnet and NWCD, not AWS directly. Contracts, support and quotas all go through them.
  • ARNs use aws-cn, endpoints differ, and a surprising number of SDK defaults assume otherwise.
  • ICP filing is a prerequisite for anything public-facing. We handle that — see ICP filing.
  • Cross-partition traffic to your global stack needs a designed path, not a VPN someone set up in an afternoon.
Capacity
Reserved, on-demand or committed-use, bought in your name at your negotiated rate. We do not resell it and take no margin on it.
Regions
Chosen against where your users are and what your residency obligations require — decided per engagement, not by default.
Our track record
We have written the reference guides on AWS China accounts, ARNs and endpoints that most people find first.

02 What we do here

BUILD

Inference on AWS China

Open-weight models running on AWS China capacity, gateway routing in front, full cost and usage records underneath, the networking and IAM to hold it together. Deployed in your account with Terraform you keep.

RUN

Operate it monthly

Capacity planning against AWS China quota and pricing, model rotations, incident response, and a monthly report that reconciles against your actual AWS China invoice.

FOUNDATION

Accounts, network, IAM

If the AWS China footprint does not exist yet, or exists as one account somebody made in 2021, we build the landing zone underneath before anything else goes on top.

03 The others

We are not loyal to any of them. The right platform is the one where your data already is, your latency is acceptable and the GPUs are actually available this quarter — and that answer changes.

Own the stack before the bill owns you.

Thirty minutes. Bring your current AI invoice. We look at what your teams send to model APIs, what it costs today, and what a self-hosted stack would cost instead. No deck, no discovery phase.