Org Intelligence Reports
Org Intelligence reports are special report types in the Aggregator that compare formal structures (roles, relationships, OKRs) with lived interaction signals from external systems.
Prerequisite: “Org Intelligence” data node switched on + at least one configured data source + person mapping for meaningful results.
Available Report Types
Section titled “Available Report Types”Engagement Score (signals_engagement)
Section titled “Engagement Score (signals_engagement)”Shows per person how well their actual interactions align with formally assigned relationships.
| Value | Meaning |
|---|---|
| > 60% | High alignment — formal role is being lived |
| 30–60% | Partial alignment |
| < 30% | Low alignment — formal role possibly inactive |
Silo Analysis (signals_silos)
Section titled “Silo Analysis (signals_silos)”Identifies formal relationships (interfaces between teams/roles) for which no interaction signals exist. High silo rate = formally defined collaboration isn’t happening.
Informal Leaders (signals_leaders)
Section titled “Informal Leaders (signals_leaders)”Detects persons who play a central role in the interaction graph (high number of strong connections) but have no formal leadership role. These persons informally coordinate and are often critical for information flow.
Formal vs. Lived (signals_vs_formal) — Org Gap Analysis
Section titled “Formal vs. Lived (signals_vs_formal) — Org Gap Analysis”Detects differences between the lived organization (actual interactions, aggregated from all configured sources) and the defined organization (structures in roleALPHA). Three gap types:
| Type | Meaning | Detection Logic |
|---|---|---|
| Undocumented | Collaboration exists in practice but is not formally defined | Aggregated signal weight person↔entity > 0.4 with no formal assignment in rA |
| Contradictory | Lived interactions contradict the formal assignment | > 60% of a person’s interactions occur outside their formal team |
| Ineffective | Defined structure in rA is not lived | Entity exists in rA, but cross-source weight < 0.1 over 90 days |
Where the Analysis Appears
Section titled “Where the Analysis Appears”- Org Gaps tab (
/signals → Org Gaps): Full overview of all gaps in the tenant — filterable by type and entity type - Entity detail panel: Direct gap analysis for a single entity (visible only to
platform_adminandtenant_admin)
Source Traceability (Anti-Hallucination)
Section titled “Source Traceability (Anti-Hallucination)”Every finding contains:
- Interaction count from real signal data (no estimation)
sourceDiversity: How many independent sources (e.g. MSGraph + Confluence + Web = 3) confirm the finding- Source URLs (for web crawler signals): direct links to the crawled pages
A finding never appears without interactionCount ≥ threshold. Descriptions always show concrete numbers, e.g. “Person appeared 14 times informally with ‘Budget Planning’”.
Multi-Source Aggregation
Section titled “Multi-Source Aggregation”The analysis evaluates all signal sources together — MSGraph, Confluence, SharePoint, Jira, CSV, web crawler. Only through combining multiple independent sources with high sourceDiversity do findings become reliable.
Thresholds
Section titled “Thresholds”These thresholds belong to the Org Gap Analysis (Org Gaps tab / entity detail panel, powered by
services/ra-signals) — not theminWeight/leaderMinEdgeCountparameters of thesignals_*aggregator reports described below under “Configurable Parameters”. Both systems evaluate signal data but are separate code paths with separate thresholds — don’t confuse the two.
| Parameter | Default |
|---|---|
| Undocumented: Min. weight | 0.4 |
| Contradictory: Outside share | > 60% |
| Ineffective: Max. weight | < 0.1 over 90 days |
| General: Min. interactions | 5 (configurable) |
OKR Alignment (signals_okr_alignment)
Section titled “OKR Alignment (signals_okr_alignment)”Shows which persons (measured by interaction strength) are most active — enabling manual comparison with OKR targets.
Creating a Signal Report
Section titled “Creating a Signal Report”- Open the Aggregator (
/aggregator). - Click New Report.
- Choose one of the signal report types (only visible when the Org Intelligence data node is switched on).
- Give it a name and save.
- Execute the report.
Configurable Parameters
Section titled “Configurable Parameters”Each signals_* report optionally reads the following fields from its config:
| Parameter | Meaning | Default |
|---|---|---|
minWeight | Minimum edge weight loaded from ra-signals | 0.1 |
leaderMinEdgeCount | Minimum edge count for a person to count as an informal leader (signals_leaders, signals_vs_formal) | 3 |
filters | Filters (same shape as the aggregator builder: field path/operator/value), applied to the computed result rows | — |
groupBy | Collapses the result rows into counts-per-group instead of passing them through individually | — |
API/MCP only: the report builder wizard currently has no editor for signal report configs (only for
builderreports). Set these parameters directly viaPUT /api/aggregator/reports/:reportIdwithconfig: { minWeight: 0.3, ... }.
Reading results
Section titled “Reading results”Every result carries, in addition to its type-specific detail fields (items/silos/leaders/persons/summary), a rows/columns pair — the same generic table/chart shape as builder reports. Use Chart type: Table for the clearest view.
Mapping quality warning
Section titled “Mapping quality warning”The more persons are mapped, the more meaningful the reports are. Persons without a mapping (no personUuid) don’t appear in the result. Check the mapping status under Person Mapping.
Related
Section titled “Related”- Aggregator — Report builder and schedules
- Org Intelligence — Privacy and how it works
- Person Mapping — Improve person mapping