Overview
The Account Brain organizes Maestro's work into departments — operational areas such as support, sales, risk, finance and knowledge. Each department runs periodic routines over the account's conversations and data and produces two main outputs: risk analysis (signals of problems or attention) and insights (opportunities and recommendations).
Each department has an autonomy level. At conservative levels, it only observes and suggests (read-only). At levels with human approval, it proposes actions that wait for your confirmation before any execution. Nothing destructive or final happens automatically without that approval.
Prerequisites
- An account with Maestro and the Account Brain enabled.
- Administrator permission to create/adjust departments and autonomy.
- Recommended: complete the onboarding (which already provisions the basic departments) and feed the knowledge corpus for richer analysis.
Step by step
- Open the Account Brain and go to the Departments area.
- Review the provisioned departments (or create/enable the ones that make sense).
- Set the autonomy level of each one (for example, read-only or with approval).
- Let the routines run: Maestro processes conversations periodically.
- Follow the risk analysis and the insights feed generated by the departments.
- When there are proposals that require approval, review and confirm (or decline) each one.
Settings & options
- Departments: areas such as support, sales, risk, finance and knowledge.
- Autonomy per department: from the most conservative (observes and suggests) to the most autonomous (executes what was approved). Start conservative and evolve.
- Human approval (HITL): proposals wait for your decision before being executed.
- Insights and risk: panels that consolidate what Maestro found, with source references when applicable. In the risk panel, a conversation's score is shown only to those who can open that conversation; each person sees the risk of the conversations they have access to.
What each insight shows
Every card in the Insight stream answers three questions, in your language (Portuguese, English or Spanish, following your user's language):
- Why this alert: the measured value and the limit of the alert, in one sentence — for example, Average first response time: 4 h 12 min — the alert starts at 4 h — and, right below it, the period it covers (Last 14 days). The period is shown only when the measurement has one: "Right now" for the current state, "Since the beginning" for cumulative indicators such as CSAT. What the card's own sentence already says is not repeated in the list of values.
- Conversations behind this alert: the conversations that contribute to the insight, each with its reason (waiting 5 h for a first reply, high risk (82/100), customer rated it 2 out of 5) and an Open conversation button. The list is built on the spot from the account's current data and shows only the conversations you have access to — exactly the ones you are able to open. Each lookup examines up to 200 conversations of the alert; when it stops at that limit, the list says there may be others. When the alert states in how many conversations the case came up and the list shows fewer, it says how many you are looking at (Showing 1 of 4 conversations).
- Suggested action: the next step, with a button that takes you to the right place — the conversation, the CRM pipeline, the charges, the knowledge base or the robot's configuration.
A recurring question insight (and the article draft suggested for it) quotes what the customer wrote. That text is shown to administrators and to anyone who has access to at least one of the conversations where the question came up. Everyone else sees in how many conversations it appeared, without the quote. When the insight points at no conversation (the article draft), the text is shown only to administrators.
Use See why and what to do to open these details. An insight that is still true is updated in the same card and moves back to the top of the list instead of becoming a repeated card. Every card shows when it was created and, when that is different, when it was last measured.
From the notification to the conversation
- The bell, the e-mail and the browser notification of an insight open that insight's card, already expanded — even when it is no longer on the first page.
- When the insight points at a single conversation, the card's main action is Open conversation, at the top of the card. With the details open, that button is not repeated on that conversation's row.
- The severity and department filters stay on screen when no insight matches the selection: the list says that nothing was found and offers Clear filters to see all of them again. While the list of a newly chosen filter is loading, Load more is disabled.
- On re-engagement insights (for example, Demand without nurturing), each conversation in the list has a Prepare follow-up button for administrators and for access profiles with the Request reply suggestions permission (to see insights, the profile also needs Manage Account Brain). The button opens the conversation and places Maestro's suggested reply in the message field. Other roles see Open conversation. You review it, adjust it and press send — nothing is sent without you. If the field already holds your own text, the suggestion is shown in a notice with an Insert into the message field button, and your text is left untouched.
- You are notified again about the same insight only when it gets worse (for example, from warning to critical). A new measurement at the same severity updates the card silently.
- Notifications about recurring-question insights do not repeat the customer's question: they say in how many conversations it appeared, and the text stays on the card, for those who may see it.
- Approval needed notifications name the department and say what is waiting for your decision.
Use cases
- Identify at-risk conversations (dissatisfaction, delays, churn) before they become problems.
- Discover sales opportunities and next-action recommendations.
- Distribute intelligence across areas, with the right level of automation for each one.
- Standardize the operation with routines that run on their own and report what matters.
Tips, limits & best practices
- Start with conservative autonomy (read-only or with approval) and increase as trust grows.
- The richer the knowledge corpus, the more accurate the risks and insights.
- Departments may appear empty at first — they populate after the routines run.
- Periodically review pending proposals so the queue doesn't pile up.
Troubleshooting
- Empty departments: confirm they are enabled and wait for the routines; check whether onboarding/ provisioning was applied.
- No insights/risk: enrich the knowledge and confirm there are enough conversations to analyze.
- Proposals don't execute: at approval levels, they depend on your confirmation — review the pending items.
- Every conversation with the same risk score: the score is a weighted sum over six signals, and a signal the account cannot produce enters as zero — indistinguishable, from the account's point of view, from "measured, and it is fine". With no SLA policy and no sentiment enrichment, four of the six drop out, the two that remain saturate within hours, and every conversation older than that lands on the same score. The panel says so when it happens — how many signals carry data and how many subjects tied — and breaks the list's ties by waiting time. To get real ordering back: apply an SLA policy, use conversation priority, or fit the saturation windows to your operation's real pace.
- The insight lists no conversations: some alerts are not about specific conversations (revenue, the CRM pipeline) and the card says so. In others the list is empty when no conversation you have access to contributes to the alert at that moment — someone with access to other inboxes may see conversations there.
- The card says "recurring question in conversations you do not have access to": the question was asked in conversations of inboxes or teams your role cannot open, so the customer's text is not shown to you. An administrator sees the full question.
- I do not see the "Prepare follow-up" button: Maestro's suggested reply is offered to administrators and to access profiles with Request reply suggestions. Ask an administrator to check your profile, or use Open conversation and write the message as usual.
- I clicked the notification and the insight "is no longer available": the insight was removed, or it is flagged as sensitive and your role has no access to it.
- "Maestro is not active in this conversation" when preparing a follow-up: the inbox of that conversation has no Maestro robot answering, so there is no suggested reply. Write the message as usual.
- An older insight shows only its text: the card says Older alert, with no details. It was recorded before Maestro started sending its measurements; new alerts of the same kind already come with the explanation.