AI Insights API
KPI anomaly detection, forecasting, action suggestions, and meeting summary generation
AI Insights API
The AI insights API provides statistical anomaly detection, time-series forecasting, AI-generated action suggestions, and meeting summary capabilities. All procedures require organization-level authentication (orgProcedure).
aiInsights.detectAnomalies
Run anomaly detection on a KPI’s time series data using Z-score analysis with exponential weighted moving average.
Input Schema:
{
kpiId: string; // KPI definition ID
windowSize?: number; // Rolling window size (default: 30)
threshold?: number; // Z-score threshold (default: 2.5)
startDate?: string; // ISO date for analysis start
endDate?: string; // ISO date for analysis end
}
Output:
Array<{
id: string;
kpiId: string;
timestamp: Date;
value: number;
expectedValue: number;
deviation: number; // Standard deviations from mean
severity: "low" | "medium" | "high";
acknowledged: boolean;
}>
aiInsights.getAnomalies
List previously detected anomalies for a KPI with pagination.
Input Schema:
{
kpiId: string;
page?: number;
limit?: number;
severity?: "low" | "medium" | "high";
acknowledged?: boolean;
}
aiInsights.acknowledgeAnomaly
Mark an anomaly as acknowledged by the current user.
Input Schema:
{
anomalyId: string;
notes?: string; // Optional acknowledgment notes
}
aiInsights.getForecast
Get time-series forecast for a KPI showing predicted values with confidence intervals.
Input Schema:
{
kpiId: string;
horizonDays?: number; // Forecast horizon (default: 30)
}
Output:
{
predictions: Array<{
date: string;
value: number;
lowerBound: number;
upperBound: number;
confidence: number; // 0-1 confidence score
}>;
}
aiInsights.getSuggestions
Get AI-generated action suggestions based on anomaly patterns and historical resolution data.
Input Schema:
{
boardId?: string; // Scope to specific board
limit?: number; // Max suggestions (default: 5)
}
aiInsights.acceptSuggestion
Convert an AI suggestion into an action item.
Input Schema:
{
suggestionId: string;
boardId: string; // Target board for the action
}
Output: Returns the created action item.
aiInsights.generateMeetingSummary
Generate an AI summary for a completed TIER meeting.
Input Schema:
{
meetingId: string;
}
Output:
{
summary: string; // Structured meeting summary
decisions: string[]; // Key decisions made
actionItems: string[]; // Action items identified
keyMetrics: Array<{ name: string; value: string; status: string }>;
}
Permissions
| Procedure | Required Role |
|---|---|
aiInsights.detectAnomalies | Any authenticated member |
aiInsights.getAnomalies | Any authenticated member |
aiInsights.acknowledgeAnomaly | Any authenticated member |
aiInsights.getForecast | Any authenticated member |
aiInsights.getSuggestions | Any authenticated member |
aiInsights.acceptSuggestion | Any authenticated member |
aiInsights.generateMeetingSummary | Any authenticated member |
Was this page helpful?