Documentation

v1.3.3.571

AI Community Manager

AI Auto-Reply#

The AI Community Manager module handles incoming comments and direct messages the instant they arrive, across every connected platform, sentiment analysis included.

AI Auto-Reply is available from the Starter plan up, with a monthly allowance of included replies per tier, 50 on Starter, 200 on Pro, 500 on Agency, unlimited on Enterprise. Past the allowance, replies keep going out and are drawn from your credit balance. On Free, incoming messages still land in the Inbox, just without an AI-generated draft or auto-send.

Comment Received Pipeline#

Processing starts the moment the platform's webhook fires, not on a polling schedule: a background task runs right after the message is saved, without making the platform wait for it. A safety-net check (every minute) only catches whatever that instant task missed.

flux

[Comment or message received via webhook]
        ↓
[Background task fires instantly, doesn't block the platform's response]
        ↓
[Keyword rule check] → match found → send instantly (skips the AI, agent hours still apply)
        ↓ no match
[Conversation context and account rules are loaded]
        ↓
[Sentiment and intent are classified, escalation keywords checked]
        ↓
[AI reply generated: GPT-OSS 20B primary, DeepSeek fallback]
        ↓
AUTO Mode   → sent after agent hours + configured delay (safety-net cron sends it if the instant path missed the window)
REVIEW Mode → draft saved + notification sent (email, and Telegram if Turbo is linked), pending approval in Inbox

Memory and conversion#

Each message is attached to a stable conversation whenever the channel provides an identifier. The engine reloads recent history, brand context, FAQs, and rules before replying. It detects product questions, purchase intent, and contact requests, then proposes a clear next step. It may voluntarily ask for a name, email, or phone number, but never invents an offer, price, or payment. Payments will require explicit activation and consent in a separate phase.

Configuration Settings (per project)#

Every setting below is scoped to the individual project (Settings → AI). A workspace with several brands configures each one independently: different mode, different hours, different tone.

auto_reply_commentsboolean

For comments: true sends automatically (AUTO), false holds a draft for manual validation (REVIEW). There's no separate on/off switch : this one field doubles as both.

auto_reply_dmsboolean

Same true/false AUTO vs REVIEW behavior as above, scoped to direct messages instead of comments.

bot_hours_start / bot_hours_end0, 23 (heure locale du projet)

Active timeframe for the agent. Outside this range: messages (and keyword-rule replies) wait until hours resume.

Auto-reply covers Facebook and Instagram comments and DMs, post comments on LinkedIn (LinkedIn private DMs are unavailable via the public LinkedIn API), and now also Email (Gmail and a professional address, see the dedicated section below). WhatsApp is intentionally excluded: Meta now offers its own native AI agent platform on that channel; TikTok does not support replies via the public API.

Email Channel (Gmail & Pro)#

A professional email inbox connects like any other channel (Settings → Accounts), and joins exactly the same pipeline described above: sync, classification, AI draft, auto/review mode, keyword rules and agent hours all apply identically. Two ways to connect: a Gmail account (Google OAuth, no password to type) or a professional address via IMAP/SMTP (host, port, credentials, works with most business email providers).

An automatic sync pass runs every 5 minutes across every connected email account (dedicated cron), no need to open the Inbox page for a new email to be picked up and processed.

Every email reply, automatic or manual, now uses a branded HTML template: logo, primary color, contact details and social links in the footer, with no unsubscribe link (a reply to one specific customer doesn't need one, unlike a bulk send). The same template dresses prospecting emails.

When a product/service catalog is configured (Settings → Brand), the reply engine cites real catalog offers and prices instead of inventing them, the same rule applies to content generation, automations and prospecting.

Outlook / Microsoft 365 is no longer offered as a dedicated integration: Microsoft has disabled standard IMAP authentication for most Outlook accounts, which made it unreliable. An Outlook/Office 365 account whose tenant still allows app-password authentication can still connect through the generic "Professional Email" form.

Keyword Reply Rules#

Beyond generic AI replies, a project can define exact-response rules: "if a comment contains this phrase, send this canned reply". Useful for a call to action like "write 'I'm interested' in the comments" answered with a link every time, word for word, with zero AI cost.

Matching is substring-based (“contains”), not exact: a rule on je suis intéressé still fires on “Bonjour, je suis intéressé !” or “je pense que je vais l'utiliser, donc je suis intéressé”.

A rule match sends instantly regardless of the project's auto/review mode (the merchant wrote the exact response, higher confidence than a generic AI draft), but agent hours and an active per-conversation pause still apply.

Configured in Settings → AI → Keyword auto-reply rules.

Per-Conversation Pause#

From any open conversation (mobile or desktop), the bell icon in the header pauses auto-reply for that one thread only: for 1 hour, 24 hours, or indefinitely until resumed. Useful when a specific customer needs a human touch without turning off auto-reply for everyone else.

Review-Mode Notification#

In REVIEW mode, the AI draft is ready the instant the message arrives, not only when the Inbox page happens to be opened. A notification goes out right away: by email always (works for every account), and additionally by Telegram if the account has an active Turbo session. A Telegram agent can only message someone who has started a conversation with it first, so email is the one channel guaranteed to reach every user.

Setting up instant auto-reply, keyword rules and notifications

YouTube

VIBE & Turbo

Website Assistant

Documentation Feedback#

Was this documentation helpful?