# Understand Your Workspace Systems

Your Workspace Systems are built in controls and monitors that keep your employees' work controlled, visible, and easier to improve.

## TL;DR

Workspace Systems are built in product systems, not employees you hire. Each owns one workspace wide job, such as spend, safety, quality, approvals, health, or reply outcomes. Open **Settings -> Technical** to see their current state and the controls available to you.

## What Is a Workspace System?

A Workspace System is a focused part of Sistava that works across your workspace, rather than for one employee. It has a fixed responsibility, records what it did, and exposes only the controls that are safe for you to change.

For a friendlier mental model, think of each system as your Officer. That language describes the role it plays for you, while the technical name in the product tells you what it actually does.

## The Seven Systems

| System                       | Think of it as your | What it does for you                                                                                     | Can you control it?                                                              | What it costs                                                                 |
| ---------------------------- | ------------------- | -------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- | ----------------------------------------------------------------------------- |
| **Rate Limits**              | Budget Officer      | Sets a workspace wide spending pace and holds new work when you reach it.                                | You choose the amount, period, and whether it is on.                             | Free local check.                                                             |
| **Cost Optimizer**           | Operations Officer  | Uses a fast lane for simple conversation and keeps real work on your selected model.                     | You can turn it on or off for the workspace.                                     | A small routing check can use credits. Savings and cost are shown separately. |
| **AI Guardrails**            | Security Officer    | Applies the safety policies you choose to every employee.                                                | You choose available policies and their settings.                                | AI safety checks use credits only while enabled.                              |
| **AI Evals**                 | Quality Officer     | Reviews selected employee work and optional reply checks against a standard.                             | You choose the workspace switch, reviewed tools, response checks, and standards. | Each enabled review is a paid AI check.                                       |
| **Human-in-the-Loop (HITL)** | Control Officer     | Makes employees ask before using tools you mark as sensitive.                                            | You choose which tools require approval.                                         | Free local policy and decision record.                                        |
| **System Health**            | Health Officer      | Shows what could block workspace progress and links you to the right next action.                        | It is system managed. You can run a check and resolve the linked issue.          | Free. Employee work that follows keeps its normal credit rules.               |
| **Conversation Monitor**     | Experience Officer  | Records your reply feedback and checks a later follow up only when it may show a reply missed the point. | Workspace policy may let an administrator turn it on or off.                     | Feedback is free. An automatic follow up review uses credits.                 |

## How to Use Them

1. Open **Settings -> Technical**.
2. Choose the system that matches the question in front of you, such as cost, safety, approval, quality, or workspace health.
3. Read the summary first, then change a control only when you want a different outcome.
4. Use the page history, activity, or log to understand what the system actually did in your workspace.

A system page is a workspace wide control panel and record. It is not a place to manage an employee.

## Why Some Systems Use Credits

Some jobs need an AI check, such as routing a message, reviewing a deliverable, checking a safety policy, or assessing a possible unresolved reply. Those checks use credits because they run additional model work beyond the employee's normal reply.

Others are deterministic product checks, such as a spending limit, approval preference, or System Health sweep. They run without an AI model and are free. Each system page explains exactly which kind it is and shows relevant cost history when there is one.

## What Systems Never Do

- They do not hire, fire, train, or impersonate an employee.
- They do not silently change your plan, buy credits, or change your billing.
- They do not turn a setting on just because another system found a problem.
- They do not make a customer facing decision for you outside the responsibility shown on their page.

## Good to Know

- Systems work across the whole workspace, so a policy or limit is not tied to only one employee unless the page says otherwise.
- A page marked **AI powered** uses an AI check for its stated job. A page marked **System** uses product logic without an AI call.
- Some controls are plan gated or policy managed. The page tells you when a control is unavailable and why.
- System Health can point you to another system when that system owns the next action. It does not change the other system's setting for you.

## Frequently Asked Questions

**Q: Are Workspace Systems AI employees?**
A: No. They are built in workspace systems with a fixed responsibility. Employees are the hires that chat, use tools, and complete your work.

**Q: Do I have to set up every system?**
A: No. Some run as part of the workspace from the start, while others remain inactive until you turn on a policy or review. Each guide explains its own default.

**Q: Can I turn every system off?**
A: No. A control exists only where turning it off is safe. System Health stays available because it is your workspace record of readiness, while other systems expose switches where you can choose the behavior.

**Q: Why does a system appear in my credit usage?**
A: It performed an AI check that required additional model work. The system page and credit usage history identify that cost separately from your employee's own work.

**Q: Which system should I open first?**
A: Open Rate Limits for spend, Cost Optimizer for model cost, AI Guardrails for content protection, AI Evals for review, Human-in-the-Loop (HITL) for approval, System Health for blockers, and Conversation Monitor for reply outcomes.