AI Intelligence
Explainable, rule-based intelligence that runs on your own server. Every insight shows the rule and the records that produced it. No external AI model is enabled.
What it does
MANARA's intelligence layer is a local rule engine, not a black box. It reads your registers and indicator values and raises insights a reviewer can verify: the rule, the threshold and the records involved are shown with every result.
Nothing leaves your server. Integration with an external AI model is not enabled in the current release; if added later, it could only be enabled by an administrator who configures it on the server and approves which data may be transferred.
Key capabilities
What works in the platform today, and what is on the roadmap.
- Available working in the current release
- Planned on the roadmap — not yet available
- Available
Indicator anomalies (z-score)
Flags values far from the indicator's recent mean and standard deviation.
- Available
Deteriorating trends
Detects consecutive moves toward a warning or alert threshold.
- Available
Treatment suggestions
For risks above appetite, suggests next steps from their treatments, linked control tests, KRIs and tolerance.
- Available
Finding clustering
Groups audit findings that share category, control and keywords.
- Available
Compliance gaps without controls
Lists requirements that have no linked control.
- Available
Repeated incidents
Highlights risks linked to several incidents.
- Available
Overdue third-party assessments
Lists suppliers due for reassessment.
- Planned
External AI model integration
Optional. Requires administrator configuration and explicit approval of data transfer.
How it connects
Records are linked across modules, so the same fact is entered once and seen wherever it matters.
See AI Intelligence with your own data
Book a guided walkthrough, or sign in to your workspace to explore the module.
Consulting: consulting@4each1.com · Projects & systems development: projects@4each1.com