AI · Coming Soon
Not available in Novigem 1.0. The AI layer is planned for a future release. Everything on this page describes what is being built. None of it is live yet. Subscribe to the changelog for release updates.
Novigem’s AI layer will transform raw Salesforce activity data into actionable intelligence using two engines that work together: one observes patterns, the other interprets what they mean.
Architecture
The separation between observation and interpretation is intentional: you will always be able to see the raw facts before any recommendations are applied.
Pattern Engine
The Pattern Engine will observe data and report what is there, finding patterns without interpreting meaning or suggesting actions.
What it will detect:
- Repetition: recurring behaviors or sequences
- Clusters: groups of similar activity or users
- Trends: changes in metrics over time
- Frequencies: how often events occur
- Correlations: relationships between variables
- Anomalies: unusual patterns or outliers
- Segment comparisons: differences between groups
Pattern outputs will be descriptive, neutral, and verifiable against your data.
Example patterns:
| Pattern | Type |
|---|---|
| ”Deals with 6+ activities per week close 68% more often” | Correlation |
| ”Follow-up within 24 hours occurs in 45% of converted leads” | Frequency |
| ”SDR team averages 12 activities per deal vs. 8 for AE team” | Segment comparison |
| ”Activity volume dropped 23% in the last 30 days” | Trend |
Insight Engine
The Insight Engine will take patterns and interpret them, explaining what they might mean and what to do about it.
What it will do:
- Infer meaning from observed patterns
- Identify risks and opportunities
- Suggest challenges, rules, or behaviors to reinforce
- Declare a confidence level for each recommendation
Example insights:
| Pattern | Insight | Confidence |
|---|---|---|
| ”Deals with 6+ activities close 68% more often" | "High activity correlates with success. Consider a challenge that rewards consistent deal engagement.” | 94% |
| “Follow-up within 24 hours occurs in 45% of converted leads" | "Fast follow-up appears to boost conversion. A rule rewarding same-day follow-up could reinforce this.” | 88% |
| “Activity volume dropped 23% in the last 30 days" | "Engagement is declining. A fresh challenge could re-energize the team.” | 72% |
Confidence levels
| Level | Range | Meaning |
|---|---|---|
| High | 85-100% | Strong pattern support, low ambiguity |
| Medium | 60-84% | Good support, some uncertainty |
| Low | Below 60% | Weak support, speculative |