Trust & Security
Enterprise AI you can control, inspect and keep.
Security teams approve AI when they can answer three questions: who is the agent acting for, what can it reach, and what did it do. UES.IO is built so the answers are always on record.
Security controls built into the platform.
Permissions
Access follows
the person.
Every agent acts with the permissions of the person it works for, down to individual rows and fields. Permissions are checked on every action, so an agent never has access its user lacks. If a person loses access, their agents lose it too.
Audit trail
Every action
on record.
Every AI call records the user, agent, model, prompt, response, tokens and cost, across all users, apps and development workspaces. When something needs explaining, the record is already there.
Budgets
Spend that stops
at the limit.
Set a budget per user. When it is reached, calls stop.
Guardrails
Guardrails per model
and per call.
Trusted and external models follow different rules, and sensitive requests get tighter controls. Combined with field-level permissions, enterprise data is not passed to an AI beyond what you allow.
Deployment
You keep control
of where it runs.
Run UES.IO hosted by us, in your own tenancy, or air-gapped and operated by you. You choose the models, including any of 14 model providers, and you can change them without rebuilding.
Configuration
Your configuration
stays yours.
Every app, workflow, agent definition and permission set exports to your own Git. You hold the record of how your agents are built and what they are allowed to do.
Built for regulated environments.
Regulated
Runs in your
own tenancy.
Supported by UES.IO. A hardened build for your home jurisdiction, with extended audit export and 24×7 support.
Sovereign
Air-gapped and
operated by you.
For environments that cannot connect out. Offline licence validation and on-site or escrowed support.
Frequently asked questions
Can an agent access data its user can’t?
No. Every agent acts with the permissions of the person it works for, down to individual rows and fields. Permissions are checked on every action.
What happens to an agent when someone loses access?
Their agents lose it too. An agent never has access its user lacks.
What does the audit trail record?
Every AI call records the user, agent, model, prompt, response, tokens and cost, across all users, apps and development workspaces.
Can we stop AI spend from running away?
Yes. You set a budget per user, and when it is reached, calls stop.
Can we choose where UES.IO runs?
Yes. It can run hosted by us, in your own tenancy, or air-gapped and operated by you. You choose the models, including any of 14 model providers, and you can change them without rebuilding.
Can we run UES.IO on our own infrastructure and control what connects to the internet?
Yes, with the Sovereign edition. You host UES.IO on premises on your own infrastructure, and you decide what connects to the internet. Sovereign also supports air-gapped operation, offline licence validation and on-site or escrowed support.