Deployment
Trio Mate supports a range of paths built around your company’s data boundary — from a quick pilot to a controlled enterprise rollout. Pick the option that best fits your team’s location, cost expectations, and compliance requirements: start small to validate, then expand step by step.
This page helps you choose among three deployment options. They don’t differ in features — they differ in where the infrastructure sits and how tightly the data path can be controlled.
1 · Three deployment options (scenario → recommendation)
Start with your situation, then match it to a recommendation:
| Your scenario | Recommended deployment |
|---|---|
| You operate in Hong Kong and want local infrastructure with a clearer data path | Hong Kong Cyberport compute |
| You want a fast start, cost control, and business validation before scaling | Alibaba Cloud Hong Kong |
| You have strict data, network, and compliance requirements and need to plan within your own environment | Private deployment |
〔Figure: scenario comparison of the three deployment options〕
Each option suits a different case:
- Hong Kong Cyberport compute — for teams that operate in Hong Kong and need local infrastructure with a clearer data path. Compute and data stay local to Hong Kong, making data flows easier to define.
- Alibaba Cloud Hong Kong — for teams that want a fast start, cost control, and business validation before scaling. It works out of the box: prove the business value first, then decide whether to move to a more controlled setup.
- Private deployment — for organizations with strict data, network, and compliance requirements. You can plan the model, compute, and system access within your own environment, keeping the boundary fully in your hands.
2 · Model and compute
Most teams don’t need to bring their own model API key: Trio already provides model access, compute, and scheduling — you just use it, with no need to separately apply for or manage each vendor’s keys.
If you have special requirements — for example, a dedicated model or a connection to custom infrastructure — you can plan for them as part of the deployment. These needs typically arise in private deployment scenarios.
3 · Next steps
Choosing a deployment option only settles “where the data lives.” What really determines “how the data is used, who can access it, and whether actions can be reviewed” is a full set of governance capabilities inside the product.
Next step: to see how these governance capabilities work together to keep things controllable, read Data & compliance →
Last updated · July 2026 ｜ Next: Data & compliance →