> ## Documentation Index
> Fetch the complete documentation index at: https://docs.exterview.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Coworkers

> How Exterview's modules combine into governed, chartered AI Coworkers: a standing role for AI on your hiring team, with a defined mandate, human-approved authority, and a full audit trail.

## Purpose

A module is a building block. An **AI Coworker** is what you actually deploy: one or more modules given a standing role and governed authority on your hiring team, the same way you'd define a job for a person. Not a raw agent, and not a chatbot — a durable role for AI, created, approved, and revoked through the same controls you'd use for a human hire.

## The Agent Charter

Every AI Coworker carries a charter. It's what makes "governed" concrete rather than a marketing word:

<CardGroup cols={2}>
  <Card title="Role & mandate" icon="id-badge">
    What the coworker is for, in one sentence a hiring manager would sign off on.
  </Card>

  <Card title="Standing goals" icon="bullseye">
    The outcomes it's accountable for across every task it runs, not just one
    session.
  </Card>

  <Card title="Scope of authority" icon="lock">
    Exactly which systems and actions it may touch. Nothing beyond an explicit
    allow-list.
  </Card>

  <Card title="Decision authority" icon="user-check">
    Where it may act on its own, and where a person must approve first.
    Human-approved by default. See [Human Review](/trust/responsible-ai).
  </Card>

  <Card title="Escalation & lifecycle" icon="user-plus">
    The named person it reports to, and a one-click suspend or revoke, the same as managing
    a human hire's access.
  </Card>
</CardGroup>

## Three coworkers, mirroring a real hiring team

Each coworker is built from modules in the [catalog](/get-started/modules) — see that page for which specific modules are available today versus upcoming. Each is separately auditable and pausable on its own, independent of the others.

<CardGroup cols={1}>
  <Card title="Sourcing & Screening" icon="magnifying-glass">
    Keeps a qualified shortlist ready for every open role: drafts and posts roles, matches
    and re-matches the candidate pool, screens resumes, and runs an initial phone screen —
    then hands the recruiter a ranked shortlist with a full audit trail.
  </Card>

  <Card title="Coordinator" icon="calendar-check">
    Keeps interview logistics and assessment moving: triages scheduling threads,
    proposes slots, and runs the avatar, coding, and panel rounds you've enabled.
    Every outgoing message and calendar action is confirmed by a person first.
  </Card>

  <Card title="Closing" icon="handshake">
    Removes offer and verification busywork: surfaces the ranked decision, runs the HR round,
    and completes verification steps. Extending an offer always stays a human decision.
  </Card>
</CardGroup>

## What this replaces

Exterview doesn't sell a headcount subtraction — it sells a supervision ratio. One recruiter oversees a portfolio of coworkers, each escalating anything outside its charter to the named person responsible for it, so oversight scales with hiring volume instead of headcount.

## FAQs

<AccordionGroup>
  <Accordion title="Is an AI Coworker a different product from the modules I already configure?">
    No. It's how the same modules are packaged, governed, and licensed together — a
    standing role built from one or more modules, not a separate agent you build yourself.
  </Accordion>

  <Accordion title="Can an AI Coworker act without a person approving?">
    Only within its charter's explicitly allow-listed scope, and consequential
    actions — anything that spends money, messages a candidate, or affects a
    hiring decision — are human-approved by default. Nothing adverse is decided
    autonomously.
  </Accordion>

  <Accordion title="Can we turn one off?">
    Yes. Suspend or revoke a coworker's charter at any time, the same way you'd
    remove a person's access, without affecting the other coworkers.
  </Accordion>

  <Accordion title="Are all three coworkers fully available today?">
    Availability follows the modules each one is built from — see the [module
    catalog](/get-started/modules) for what's live versus upcoming. The charter and
    governance model apply identically regardless of which modules are enabled yet.
  </Accordion>
</AccordionGroup>

## Related

<CardGroup cols={2}>
  <Card title="Modules" icon="grid-2" href="/get-started/modules">
    The individual agents each coworker is composed from.
  </Card>

  <Card title="Smaya" icon="sparkles" href="/smaya/overview">
    The companion that helps you ask about and direct your coworkers' work.
  </Card>

  <Card title="Responsible AI" icon="scale-balanced" href="/trust/responsible-ai">
    The human-review standard every charter is built on.
  </Card>

  <Card title="Core Concepts" icon="book-open" href="/get-started/core-concepts">
    The platform's mental model and vocabulary.
  </Card>
</CardGroup>