Behavioral Insights Engine

Advanced analytics and anomaly detection for your Discord server's community health

👑 ProPro tier feature

Key Features

Behavioral Analysis

👑 Pro

Comprehensive tracking of messages, voice activity, member joins, reactions, and engagement patterns across your server.

Health Scoring

👑 Pro

Automatic health scores (0-100) based on activity consistency, engagement stability, and absence of anomalies.

Anomaly Detection

👑 Pro

Real-time detection of unusual patterns including spikes, drops, engagement cliffs, and behavioral deviations.

Incident Alerts

👑 Pro

Automated notifications for significant behavioral anomalies with severity levels and correlation details.

Activity Heatmaps

👑 Pro

Visual representations of activity patterns by day and hour to identify peak engagement times and trends.

Data Privacy

👑 Pro

End-to-end data privacy with configurable retention, exclusions, and full GDPR/CCPA compliance.

What BIE Tracks

Behavioral Metrics

  • •
    Message Activity

    Count, frequency, timing, and channel distribution of messages

  • •
    Voice Activity

    Voice channel joins, duration, participants, and peak times

  • •
    Member Engagement

    Reaction frequency, thread participation, and interaction patterns

  • •
    Membership Changes

    Member joins, leaves, and role changes with timing analysis

  • •
    Temporal Patterns

    Activity distribution across hours, days, and weeks

What BIE Does NOT Track

BIE analyzes only metadata and behavioral patterns. It never stores message content, user identities beyond IDs, or personal information. Privacy is built into every analysis.

Health Scoring System

The health score is a 0-100 metric that reflects your server's community vitality and engagement stability. Higher scores indicate a healthy, active, and engaged community.

Score Interpretation

85-100
Excellent

High engagement, stable activity, no anomalies

70-84
Good

Solid engagement with minor fluctuations

50-69
Fair

Moderate engagement with notable variations

30-49
Poor

Low engagement or frequent anomalies detected

0-29
Critical

Significant issues requiring immediate attention

Scoring Factors

  • →
    Activity Consistency (30%)

    Stability of daily and weekly activity patterns

  • →
    Engagement Levels (25%)

    Message, voice, and reaction frequency

  • →
    Member Participation (20%)

    Percentage of active members and repeat participants

  • →
    Anomaly Absence (15%)

    Lack of severe anomalies or unexpected patterns

  • →
    Trend Direction (10%)

    7-day and 30-day trend momentum

Improving Your Health Score

Focus on consistent engagement, encourage member participation, and maintain regular activity patterns. Investigate and resolve alerts to prevent anomalies from impacting your score.

Anomaly Types

BIE detects several categories of behavioral anomalies. Understanding these helps you respond appropriately to community changes.

Activity Drop

Sudden significant decrease in overall server activity, messaging, or voice participation compared to baseline patterns.

Activity Spike

Unusual surge in messages, voice activity, or member joins beyond expected patterns for the time period.

Engagement Cliff

Rapid decline in member interaction, reactions, and participation despite messages still being posted.

Mass Membership Change

Unusual number of members joining or leaving within a short timeframe, often indicating external events or raids.

Pattern Deviation

Activity occurring at unusual times or from unusual channels, breaking established temporal or spatial patterns.

Voice Anomaly

Unusual voice channel activity, duration, or participant patterns compared to historical baselines.

Severity Levels

Each anomaly is assigned a severity level based on the magnitude of deviation from baseline and potential impact on community health.

CRITICAL

Severe deviation (>50% change) with significant community impact. Immediate investigation recommended.

HIGH

Major deviation (30-50% change) with notable community effects. Close monitoring and response suggested.

MEDIUM

Moderate deviation (15-30% change) warranting attention. Keep an eye on trends and correlate with other events.

LOW

Minor deviation (<15% change) from baseline. Typically not actionable alone but useful for trend analysis.

Data Privacy & Compliance

SYNTHET takes data privacy seriously. BIE is designed with privacy-first principles and full compliance with international regulations.

What We Store

  • • Activity timestamps
  • • Event counts and frequencies
  • • Channel and member IDs
  • • Behavioral patterns
  • • Aggregate statistics

What We Never Store

  • • Message content
  • • User names or usernames
  • • Personal information
  • • Voice transcripts
  • • File contents

GDPR Compliance

Full compliance with data subject rights including access, rectification, and erasure. Configure retention policies and exclusions to manage your data.

CCPA Compliance

California residents have full rights to access, delete, and opt-out of data processing. All data handling is transparent and user-controlled.

Data Retention Policy

Configure your retention settings to automatically purge data after 30 days to 2 years. Manual purges are available anytime. Longer retention enables better trend analysis but increases data storage.

Frequently Asked Questions

What is the Behavioral Insights Engine (BIE)?
The Behavioral Insights Engine is a Pro-tier feature that analyzes your server's activity patterns, detects anomalies, and provides actionable insights through health scoring and incident alerts. It tracks behavioral metrics across your entire community.
How does health scoring work?
Health scoring uses a 0-100 scale based on activity consistency, engagement patterns, member participation, and absence of anomalies. Higher scores indicate stable, healthy community engagement. The score updates daily and factors in rolling 7-day and 30-day trends.
What types of anomalies does BIE detect?
BIE detects activity drops, sudden spikes, engagement cliffs, mass member joins/leaves, unusual message patterns, voice activity deviations, and reaction frequency changes. Each anomaly is categorized by severity.
Is my data private with BIE enabled?
Yes. BIE only analyzes metadata (timestamps, counts, patterns) and never stores message content. Data is retained according to your configured retention policy and can be purged manually. Full GDPR and CCPA compliance is maintained.
Can I exclude channels or members from analysis?
Yes. You can exclude specific channels (like announcements or bot-spam) and members from behavioral analysis to focus on relevant community engagement. Exclusions are configurable per channel and member.
How often is data analyzed?
BIE analyzes activity in real-time as events occur, with daily health score calculations and incident detection running continuously. Historical data is available for review at hourly, daily, weekly, and monthly granularities.