Deployment
Customer-hosted, with your own AI model.
Each customer runs a dedicated AxenticOne installation in its own cloud account or on its own servers. TechXio installs and supports it but does not host it. Your data stays in your environment and is never shared between installations.
Where it runs
One installation per company, inside your boundary.
An installation serves one operating company with one accounting system. It is packaged as containers and runs with its datastores on a private network, behind your own HTTPS front end.
- Your cloud account or your own servers
- Datastores on a private network, never exposed publicly
- Backups and restores with tools that ship with the product
- All data stays in your environment and can be exported in full
Your boundary. Everything starts inside your own cloud account or on your own servers.
1/6Read every step
- Your boundary. Everything starts inside your own cloud account or on your own servers.
- Your installation. One AxenticOne installation serves your company. TechXio installs it but does not host it.
- Your AI model. Agents use a model account you own. You pay your model provider directly.
- Your data. Orders, documents, revisions and the audit trail stay in your environment and can be exported in full.
- Your connections. The installation connects to your mailboxes, supplier channels and accounting system.
- Nothing shared. Another company’s installation is separate. Data is never shared between installations.
Your AI model
Agents use a model account that you own.
You connect a commercial model service under your own account, and pay your provider directly. Support for locally hosted models, and a validation suite that checks quality on your chosen model before go-live and after any change, are on the roadmap.
Each capability is labelled:Available in the product todayOn the roadmap specified and planned
Single-company installation on your infrastructure
AvailableA Docker Compose installation in your cloud account or on your servers. Datastores stay on a private network.
AI model under your own account
AvailableConnect a commercial AI model service under your own account and pay it directly. Google Gemini is supported today.
Choice of model, including locally hosted
On the roadmapAn administrator selects the model for each agent from those you make available, including models you host yourself.
Test mode
On the roadmapProcess orders end to end without sending anything to external parties or posting to the accounting system.
Configuration export and import
On the roadmapMove the full configuration of an installation to another, for example from a test environment to production.
Responsibilities
Who does what.
You own the infrastructure and the model account. TechXio installs, validates and maintains the product.
| Responsibility | You | TechXio |
|---|---|---|
| Infrastructure and its cost | Yes | No |
| AI model account and its cost | Yes | No |
| Installation and model validation at setup | Provides access | Yes, under the setup fee |
| Servers, backups and monitoring | Yes | Guidance and tools |
| Maintenance releases, fixes and support | Applies releases | Yes, under the subscription |
| A model change after go-live | Requests and approves | Runs the validation before it goes live |
Talk to your IT lead with us.
We will share the reference environment, the minimum requirements and how the installation is secured.