AI Insights

AI Insights continuously analyzes your workspace data to surface opportunities, flag issues, and provide actionable recommendations.

Where Insights Appear

Dashboard Widget

The primary location for AI Insights is the Dashboard:

• Scroll to the AI Insights section
• View up to 5 recent insights
• Click to expand or dismiss

Full Insights Page

For a complete feed:

1. Navigate to /insights or click View All in the widget
2. See all insights from the last 30 days
3. Filter by type, severity, or module

Email Digests

Important insights can be included in your email digests (if enabled in notification settings).

Types of Insights

AI Insights are categorized by type:

The 9 Pattern Rules

The AI engine monitors for 9 specific pattern types:

1. Revenue Variance

Detects unusual changes in revenue trends.

Triggers:

• Significant increase or decrease (±15% threshold)
• Compared to previous period (week/month/quarter)

Example:

"Revenue variance detected: $45,000 this week vs $52,000 last week (-13.5%)"

2. Expense Anomaly

Identifies unexpected expense patterns.

Triggers:

• Category spending 2x+ above average
• Unusual expense timing or amount

Example:

"Expense anomaly: 'Travel' category is 3x higher than your 90-day average"

3. Support Ticket Spike

Alerts on sudden increases in support volume.

Triggers:

• 40%+ increase in ticket volume vs previous period
• Pattern sustained over 24+ hours

Example:

"Ticket spike alert: 28 new tickets in 24 hours (avg: 12)"

4. Pipeline Bottleneck

Identifies deals stuck in pipeline stages.

Triggers:

• 5+ deals in a stage for 30+ days
• Stage duration 2x+ longer than average

Example:

"Pipeline bottleneck: 6 deals stuck in 'Proposal' for over 30 days"

5. Recruiting Funnel Analysis

Highlights candidate drop-off points.

Triggers:

• Stage-to-stage conversion drops below 50%
• Unusual time-in-stage patterns

Example:

"Funnel insight: 60% of candidates drop off between 'Screening' and 'Interview'"

6. Overdue Task Detection

Surfaces tasks past their due date.

Triggers:

• Tasks overdue by 1+ days
• Assigned to you or your team

Example:

"You have 7 overdue tasks. 3 are high priority."

7. SLA Breach Warning

Warns of approaching support SLA violations.

Triggers:

• Ticket approaching resolution time limit
• Urgent ticket unassigned for 1+ hours

Example:

"SLA warning: 2 tickets approaching 4-hour resolution target"

8. Deal Stagnation

Flags opportunities with no recent activity.

Triggers:

• High-value deal (> $10K) no activity for 14+ days
• Any deal no activity for 30+ days

Example:

"Deal stagnation: 3 deals worth $85K total have no activity in 14 days"

9. Cross-Module Correlation

Finds connections between different modules.

Triggers:

• Statistical correlation with p < 0.05
• Actionable insight from pattern

Example:

"Cross-module insight: Closed deals correlate with reduced support tickets from those customers"

Interacting with Insights

Dismissing Insights

If an insight isn't relevant:

1. Click the × or Dismiss button
2. The insight is removed from your feed
3. Similar insights may still appear in the future

Acting on Insights

Most insights include a suggested action:

1. Review the insight
2. Click View Details to see more context
3. Click Take Action to go directly to the relevant page
4. Or click Dismiss if no action is needed

Snoozing Insights

To temporarily hide an insight:

1. Click the ... menu on the insight
2. Select Snooze
3. Choose duration: 1 day, 1 week, or 1 month
4. The insight reappears after the snooze period

Insight Severity Levels

Insights are color-coded by importance:

Configuring Insights

Enabling/Disabling Patterns

Control which insight types you receive:

1. Go to Settings > AI Configuration
2. Toggle Proactive Suggestions on/off
3. Or disable specific pattern types

Notification Preferences

Choose how you receive insights:

1. Go to Settings > Notifications
2. Configure AI Insights settings:
   • In-app only
   • Include in email digest
   • Real-time notifications for critical insights

Sensitivity Tuning

Adjust how sensitive the AI is to patterns:

To change: Settings > AI Configuration > Insight Sensitivity

Insight Accuracy

Confidence Scores

Each insight includes a confidence indicator:

• High confidence (> 90%) — Strong pattern, reliable
• Medium confidence (70-90%) — Likely valid, worth checking
• Low confidence (< 70%) — Possible pattern, verify before acting

When to Trust Insights

Reporting Incorrect Insights

If an insight seems wrong:

1. Click the ... menu on the insight
2. Select Report Issue
3. Describe why it's incorrect
4. Submit

This helps improve the AI's accuracy over time.

Insight History

Viewing Past Insights

1. Go to /insights (full insights page)
2. Filter by date range
3. See dismissed insights by checking Show Dismissed

Insight Trends

Over time, track:

• Most common insight types
• Resolution rates
• Pattern frequency changes

Best Practices

Responding to Insights

1. Critical/Warning: Address promptly
2. Info: Review during regular planning
3. Insight: Consider for strategic decisions
4. False positives: Dismiss and report

Using Insights for Planning

Weekly review of insights can inform:

• Pipeline reviews (deal stagnation alerts)
• Support capacity planning (ticket spikes)
• Hiring decisions (funnel analysis)
• Budget adjustments (expense anomalies)

---

Next Steps

• AI Configuration — Tune insight settings
• Dashboard — See insights in context
• AI Assistant — Ask questions about your data