# The workspace for governed AI.

> Synainesi Desktop for everyday AI work, Guard for enforcement on the device and Cloud for governance: AI your people want, within rules you set.

Canonical: https://synainesi.com/product/

Three parts

## One set of rules.

Synainesi separates the place where people work, the place where rules are enforced and the place where rules are decided. That is what lets AI move fast while control stays firm.

Desktop

### Where the work happens

A calm, capable AI workspace for Windows and macOS. Documents, research and everyday tasks, with confidential work kept on the device.

Guard

### Where rules are kept

A local enforcement layer that every model and tool request crosses. It checks what would leave, applies your policy and keeps a clear record of each decision.

Cloud

### Where rules are decided

One control plane for approved models, entitlements, budgets and oversight. It sees how AI is used, not what people are working on.

Product applications  
Desktop-first. Cloud controlled.

## Turn complex work into trusted progress.

A calm workspace your people want to use, with protection they can see, and one control plane where your organisation sets the rules.

Desktop app — Home / workspace

AI work screen — with protection indicators

Cloud control plane — Admin / governance

Concept designs shown with sample data.

How it flows

## Every request crosses one boundary.

### Ask

Your people work with the models and agents they are allowed to use.

### Check

Policy is applied on the device, and sensitive values are protected before anything leaves.

### Approve

Consequential actions wait for a person, and approval covers exactly what was approved.

### Prove

Each decision leaves evidence, so oversight never depends on collecting everyone’s work.

Agents

## Agents have permissions, not trust.

Synainesi agents do the reading, comparing and organising that regulated work demands, within permissions your organisation sets. People make the decisions that matter.

Designed for regulated work: regulatory monitoring, policy review and KYC file review, each with a clear line it will not cross.

[Read: Agents have permissions, not trust](https://synainesi.com/insights/agents-have-permissions-not-trust/)

---

Synainesi: AI within your rules.

Source: https://synainesi.com/product/. The HTML page is canonical; this Markdown version is provided as a convenience.
