# Things that control how they behave

Just ask. It configures itself.

Skip the settings panels. Tell your AI employee what skills to learn, what rules to follow, and what personality to use, all through natural conversation. They can even browse the skill catalog and configure themselves as work evolves.

Most AI tools force you into settings panels and dropdown menus. This one listens. Tell your employee what to do in plain language, and it configures its own skills, duties, persona, and AI model through conversation.

Skills control capabilities, duties enforce rules, personas shape personality, and model selection balances cost against performance. All are managed through chat.

Your employees can even configure themselves and the platform, adjusting their own skills and settings as work evolves.

Three layers decide how an employee behaves, and they stay separate on purpose. Skills are what it is good at, duties are the rules it never breaks, and the persona is how it sounds. Change one and the other two are untouched, so tightening a rule cannot quietly rewrite the tone, and rewriting the tone cannot drop a rule. Your employee then picks the right skill for each task on its own, which means you describe the work rather than the method.

You also decide how hard an employee thinks and how far it is allowed to go. Thinking depth runs from off through minimal, light, medium and deep, or you leave it on auto and Sistava picks the right depth per message. A separate limit sets how many actions one employee may take on a single task before it stops and reports back, anywhere from 50 up to 1,000. A research role gets room to search, read and delegate, while a simple support role stays tight, and both settings apply across the workspace with a per-employee override when one role genuinely needs something different.

Credits Saver is on by default and quietly keeps everyday chat cheap. Greetings, quick acknowledgements and small edits like make that shorter are handled by a lighter model, while anything that counts as real work still runs on the model you chose for that employee. On ordinary back and forth that often means 40 to 50 percent fewer credits with no drop in quality where it counts. It is per employee and fully in your hands, so you can switch it off whenever you want.

Nothing here is locked in. Every field on an employee's Profile tab, including name, role, team, avatar, persona, language and tone of voice, is editable in place, and the change applies from its next message onward across every channel that employee works on. A change history on the same tab records every edit, including the ones the employee made to itself, so you can always see what changed and put it back. Tell your team leader once how you want to be addressed, when you work and how you like answers written, and every employee on the team picks that up without a form to fill in.

## At a Glance

- **0** Settings pages needed
- **Chat** Configure through conversation
- **Auto** Self-adjusting skills

## Benefits

### Conversational setup

Say 'learn advanced SEO research' and the agent activates the right skill. Say 'always check with me before sending emails' and it creates a duty. No settings panels, no admin overhead.

### Rules that stick

Duties are non-negotiable behavioral rules your AI employee follows on every interaction. They persist across all conversations and all users. A duty is the difference between a one-time request and a standing policy.

### Self-managing agents

Your AI agents can browse the skill catalog, activate new capabilities, and adjust their own configuration as the job demands change. You define the boundaries, they operate within them.

## How It Works

**You shape how your AI employees think and act by assigning skills they master, duties they always follow, and a persona that fits your brand.**

Skills define what an AI agent is good at: writing, research, data analysis, customer communication, and dozens more. Duties are behavioral rules the agent follows unconditionally, things like always escalating refunds over a certain amount, never making promises about delivery dates, or always responding in a formal tone. Together, skills and duties give each agent a precise and reliable character.

You also choose the underlying AI model for each employee, matching the right capability level to the right role. A lightweight model handles routine support tickets, a more powerful one tackles complex research. Agents self-manage within their defined scope, deciding which tools to use and how to approach a task, without needing you to supervise every step.

Self-configuration has edges, and you draw them. An employee can browse the skill catalog, activate what a job needs and adjust its own settings, but the ability to operate the platform itself is a capability you hand out deliberately rather than something every employee has, and anything that touches billing always waits for a person. Every change is recorded with a timestamp and the reasoning the employee gave, so you can review it later or reverse it from the change history. You set the boundaries once, then let them work inside those boundaries.

## Use Cases

### Assign skills that define what they're great at

Give your employee specific skills like negotiation, technical writing, or data analysis. Skills shape how they approach problems and what methods they apply.

### Set duties they follow every time

Duties are non-negotiable rules: always escalate billing issues, never make promises about timelines, include a summary at the end of every report. Consistent behavior, no reminders needed.

### Give them a persona that fits your brand

Define their name, tone, communication style, and personality. A customer-facing employee sounds warm and professional. An internal analyst is direct and concise.

### Switch models to match the task

Run fast, cheap models for high-volume routine tasks and switch to more powerful models when the work demands deeper reasoning. One employee, the right model per job.

## FAQ

### How do I control the way an AI employee behaves?

You control an AI employee's behavior with three things: skills, duties and a persona. Skills are the capabilities it brings to a job, duties are standing rules it follows on every task, and the persona sets its name, tone and communication style. In Sistava you set all three by saying what you want in ordinary conversation, and the employee configures itself to match.

### Do I need to write prompts or learn a settings panel for this?

No. Configuration in Sistava happens in chat. Say learn advanced SEO research and the employee finds and activates that skill; say always check with me before sending an email and it turns that into a standing rule. Every setting also has a normal screen if you would rather click through it, but most people never open one.

### Skills, duties and a persona: which one should I use for what?

Use a skill when you want an employee to be good at something, such as writing case studies or qualifying leads. Use a duty when you want a rule that holds on every single task, such as always summarising at the end or never quoting a delivery date. Use the persona when you want to change how it sounds rather than what it does. The three are independent, so changing one leaves the other two exactly as they were.

### How quickly does a change take effect?

Immediately. A new skill, a new duty or an edited persona applies from that employee's next message onward, on every channel it works on, with no rehire, redeploy or restart. Fields on the Profile tab save the moment you click away, so there is no separate save step to forget.

### Can I stop an employee from changing its own configuration?

Yes. Operating the platform itself is a capability you assign deliberately, usually to a team leader, rather than something every employee has. Anything that affects billing always needs a person to approve it. Every change an employee makes is logged with a timestamp and the reasoning it gave, and you can reverse it from the change history.

### What if I set something up wrong and want to undo it?

Every edit lands in a change history on that employee's Profile tab, including edits the employee made to itself, so you can see exactly what changed and when. Skills can be deactivated, duties removed and persona text rewritten at any time. Nothing here is permanent and nothing needs support to reverse.

### Do I have to configure every employee one by one?

Skills and duties are set per employee, which is exactly what keeps a support role and a research role behaving differently. What you never repeat is yourself: tell your team leader once how you want to be addressed, when you work and how you like replies written, and every employee on that team picks it up. Skills are reusable too, so one custom skill assigned across the team keeps everyone consistent, and updating it updates all of them.

### Can I make an employee think harder on complex work and lighter on simple work?

Yes, and there are two dials for it. Thinking depth runs from off through minimal, light, medium and deep, or auto lets Sistava choose per message. Separately, Credits Saver sends greetings and small edits to a lighter model while real work stays on the model you selected, so the easy turns cost less without touching the quality of actual output.

## Agents That Follow The Rules

Giving an AI agent the right skills is only half the job. The other half is making sure it behaves the way your business requires. Duties lock in the non-negotiables: compliance requirements, communication standards, escalation triggers, and brand voice. Your agent executes autonomously inside those guardrails, not around them.

When your business needs change, you update the behavior in one place and every affected agent adjusts immediately. No retraining, no re-briefing, no hoping the agent remembers. You define the rules once and trust that they stick.

## How It Works

1. **Assign Skills And Duties**: Give each AI employee the capabilities they need and the behavioral rules they must always follow, no exceptions.
2. **Set The Persona**: Define tone, name, and communication style so the agent represents your brand consistently across every interaction.
3. **Choose The Right Model**: Match the underlying AI model to the complexity of the role, from fast and efficient to deep and analytical.

> Setting duties was a game changer. My sales agent stopped overpromising on timelines the moment I added that as a rule. Zero micromanagement needed.
>
> Marcus T., Founder · B2B startup

## Features in Behavior

- [Shape Your Employee's Personality](/en/features/behavior/employee_persona): Write a short persona prompt to set how each AI employee talks: formal or casual, brief or detailed, warm or blunt. Every employee starts with a role-based default, and you can rewrite it any time from the Profile tab. The change applies to the employee's next message, no restart needed.
- [Customize Employee Profile](/en/features/behavior/employee_profile_customization): Every hire's Profile tab is fully editable: name, role, team, avatar, persona, language, and tone of voice, all changed in place without a rehire. Updates apply from the next message onward, across every channel the employee uses. A change history on the same tab shows who edited what and when, including edits the employee made to its own persona.
- [Cost Optimizer](/en/features/behavior/credits_saver): Cost Optimizer is the built-in Credits Saver model router. It watches every message and quietly hands the easy ones (greetings, quick acknowledgements, 'make that shorter') to a lighter, cheaper model, while real work still runs on the model you picked. On everyday back-and-forth that often means 40 to 50% fewer credits, with no drop in quality where it counts. It is on by default and fully in your control: turn it on or off anytime, and see tracked credits saved, its router cost, and the net benefit on its own page.
- [Ask Your Employee About Sistava](/en/features/behavior/platform_awareness): Ask your employee about Sistava itself and get a real answer: what plan you're on, what a feature costs to unlock, what your limits are, or who else you could hire. Your employee reads the live plan and feature data at the moment you ask, so the answer matches what your account actually has right now, not a guess from training data.
- [Choose How Deeply Your Employee Thinks](/en/features/behavior/reasoning_effort_control): Set how much an employee thinks before it answers, from Off through Minimal, Light, Medium, and Deep, or leave it on Auto so the platform picks the right depth per message on its own. It applies workspace-wide from the AI Behavior settings page, with a per-employee override when one employee's work genuinely needs a different depth than the rest of the team.
- [Set a Per-Task Action Limit](/en/features/behavior/max_actions_override): Set exactly how many actions one employee can take while working on a single task, from 50 up to 1,000. Give a research-heavy employee more room to search, read, and delegate before it has to stop and report back, or cap a simple support employee tighter so it never wanders past what the task needs.
- [Skills Catalog](/en/features/behavior/skills_catalog): Give any AI employee a step-by-step playbook for a specific type of work: how you want blog posts researched, how outreach emails get structured, how a report should be formatted. The employee reads the full skill only when a task actually matches it, so you can attach as many as you want without slowing anything down.
- [Duties Catalog](/en/features/behavior/duties_catalog): Give any employee a set of standing rules they follow on every single interaction, no exceptions: respond within the hour, always CC you on sales emails, never discuss pricing without approval. Pick from the built-in catalog or write your own in plain English, and the employee carries the full text of every active duty in mind on every turn.
- [Personal Preferences Memory](/en/features/behavior/founder_personal_profile): Tell your team leader once how you want to be addressed, when you're working, and how you like answers written, and every employee on your team picks it up automatically. No settings form to fill out: just say it in chat and it sticks.
- [Skills They Can Master](/en/features/behavior/custom_skills): Teach any AI employee your own step-by-step process, from a blog-writing checklist to a lead-qualification script, and it follows those exact steps whenever a matching request comes in. Write it yourself in the Skills tab, or just describe your process in chat and the employee turns it into a skill. Skills load only when the task matches, so you can add as many as you want without slowing anything down.

## Explore

- [Every feature](/en/features)
- [Browse AI employees](/en/agents)
- [Pricing](/en/pricing)