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

# Management

> Overviews, nudges, and the pulse of the operation for whoever leads — without micromanaging. This is where the Digital Twin shines.

It's 7:17 a.m. Before the first coffee, the phone already has eleven
messages: "did the delivery project ship?", "did we close that client?",
"did the board move?", "and the monthly target?". Each one means opening a
different screen, recalling a detail, replying. None is hard. Together they
eat your morning — and leading turns into repeating status.

**What leaves your plate:** the effort of *digging up* the state of the
operation and of *chasing* what's stalled. You still decide what matters —
but you arrive at the decision with the picture already assembled, instead
of having to build it.

<Note>
  Leading isn't knowing everything every minute. It's being told **at the
  right moment** what changed and what stalled — and being able to trust
  that the rest is moving. That "pulse" is exactly what the agents keep for
  you.
</Note>

## The pattern, applied to management

Every flow below follows the same shape: a **trigger** fires → the agent
pulls the right context from the [Company Brain](/en/features/brain) →
decides and uses real tools → anything **sensitive it proposes** and waits
for you → everything **becomes memory** and improves the next cycle.

```mermaid theme={null}
flowchart LR
    A[Trigger<br/>end of day, question, deadline, risk] --> B[Pull context<br/>boards, CRM, finance, Brain]
    B --> C[Assemble the picture<br/>and decide what to do]
    C --> D{Touches people<br/>or the outside?}
    D -->|No| E[Notify and record]
    D -->|Yes| F[Propose + wait for you]
    E --> G[Becomes memory]
    F --> G
```

## Four concrete flows

### 1. The daily (and weekly) digest, unprompted

Every Friday at 6 p.m., or every end of day, a [routine](/en/features/rotinas)
wakes the [Chief of Staff](/en/agents/athena). She sweeps what changed on
the [boards](/en/features/boards), in the [CRM](/en/features/crm), and in
[finance](/en/features/financeiro), cross-checks it against what the Brain
expected, and assembles a short overview: **what moved, what stalled, what's
due tomorrow**. It lands in your **notifications inbox** — the bell the
agents ring on their own — and, if you want, becomes a
[document](/en/features/documentos) you skim in thirty seconds.

> It's not a dashboard you have to go look at. It's a heads-up that arrives
> — with what changed up top and what's stuck already pointed out.

### 2. The Twin answers "how's X going?" in your voice

The same status questions land all day, and the answer is almost always on
a board or a lead. Your [Digital Twin](/en/agents/digital-twins) — which
speaks in **your voice** — handles the repetitive ping: it reads the real
state in the [CRM](/en/features/crm) or the [boards](/en/features/boards),
checks it against the Brain, and replies in [chat](/en/features/chats) or on
[WhatsApp](/en/integrations/whatsapp) the way you would. Simple status it
resolves; when the question turns into a **decision or a commitment**, it
stops and pings you.

<Info>
  The Twin answers *status facts* in your voice. It does **not** take on a
  new commitment, promise a deadline, or say "yes" for you on anything that
  binds the company — that it brings back for you to decide.
</Info>

### 3. The nudge you don't have to send

A task is past due. The Chief of Staff notices — because she tracks the
[boards](/en/features/boards) — and prepares the **nudge** to the owner. If
the owner is on the team, she sends a gentle internal reminder and records
the thread on the task itself. If the nudge has to go **outside** (a vendor,
a partner) over email or WhatsApp, she **drafts and proposes** it — you read
it, adjust the tone, approve the send. You stop being the bottleneck for
chasing work without losing control of how the company speaks.

### 4. The heads-up before it becomes a problem

The best time to learn a target will slip is **before** it slips. A periodic
[routine](/en/features/rotinas) keeps an agent watching the numbers that
matter — an entry in [finance](/en/features/financeiro), the pace of the
funnel in the [CRM](/en/features/crm), a critical deadline on a board. When
a reading drifts out of the expected range, it **raises its hand** in your
notifications inbox with the context: what changed, since when, and what it
suggests you look at. And before a meeting, that same pattern assembles a
**brief** from the Brain — a summary of what matters about that client or
project, ready in a document, so you walk in informed instead of improvising.

| Trigger             | The agent does on its own          | The agent only proposes              |
| ------------------- | ---------------------------------- | ------------------------------------ |
| End of day / week   | Assembles and delivers the digest  | —                                    |
| "How's X going?"    | Answers the status in your voice   | Taking on a new commitment           |
| Overdue task (team) | Internal reminder + record         | —                                    |
| External nudge      | Drafts the text                    | **Sending** it to the vendor/partner |
| Metric out of range | Flags it with context              | Any action touching money/contracts  |
| Before a meeting    | Assembles the brief from the Brain | —                                    |

## What stays human

<Warning>
  **Management has decisions an agent never makes alone.** Here it only
  **proposes** — and waits for your sign-off:

  * **People decisions** — hiring, promoting, letting go, reshuffling roles.
    The agent gathers context; the judgment is yours.
  * **Performance judgments** — who's doing well, who needs support. It shows
    the surface of work that moved, without judging the person.
  * **Any external commitment** — a promised deadline, a proposal to a client,
    an agreement with a vendor.
  * **Sensitive comms** — a hard nudge, a delicate announcement: the Twin
    **drafts in your voice and proposes**, you decide if it goes out.
  * And the universals: **money** (payment/refund), **contract/signature**,
    **regulatory or legal communication**, **governance decisions**, and
    **destructive actions**.

  All of it is kept in an audit trail — you see what was proposed, why, and
  what you approved.
</Warning>

## Where to start

<Steps>
  <Step title="Turn on the end-of-day digest">
    Ask the Chief of Staff for a daily or weekly [routine](/en/features/rotinas)
    that delivers "what moved and what stalled". That's the overview coming
    to you — the smallest first step in management.
  </Step>

  <Step title="Let the Twin handle status pings">
    Set up your [Digital Twin](/en/agents/digital-twins) with a few examples
    of your voice and hand it the repetitive status questions. Anything that
    becomes a decision it gives back to you.
  </Step>

  <Step title="Raise autonomy as it gets it right">
    Start with everything proposed for you to approve. As the internal nudges
    and the digest prove accurate, release more. Trust is a ratchet, not a
    leap.
  </Step>
</Steps>

## Next steps

<CardGroup cols={2}>
  <Card title="Digital Twin" icon="fingerprint" href="/en/agents/digital-twins">
    The twin that answers in your voice and keeps the pulse of the operation
    for you.
  </Card>

  <Card title="Chief of Staff" icon="user-tie" href="/en/agents/athena">
    The org's operator who assembles the digest, chases deadlines, and flags
    risks.
  </Card>

  <Card title="Routines" icon="clock-rotate-left" href="/en/features/rotinas">
    The scheduling behind the daily digest and the metric monitoring.
  </Card>

  <Card title="Planning" icon="map" href="/en/use-cases/planning">
    The upstream of management: goals become tasks and priorities get clear.
  </Card>
</CardGroup>