Skip to content

Insights & Analytics

CalcBridge provides AI-powered insights that analyze nominated workbook data to surface patterns, anomalies, and actionable recommendations.

Overview

The insights system:

  • Nomination Workflow: Select which workbooks to analyze
  • Stakeholder Views: Tailor insights for executives, finance, operations, or analysts
  • AI Generation: Automated pattern detection and anomaly analysis
  • Natural Language Q&A: Ask follow-up questions about any insight
  • Session Management: Track insight history with regeneration support
flowchart TD
    NOM["Nominate\nWorkbooks"] --> EXTRACT["Extract\nCanonical Data"]
    EXTRACT --> AI["AI Analysis\n(SecureAIGateway)"]
    AI --> INSIGHTS["Generated\nInsights"]
    INSIGHTS --> QA["Natural Language\nQ&A"]

    subgraph Categories
        FH["Financial\nHealth"]
        OE["Operational\nEfficiency"]
        RC["Risk &\nCompliance"]
        TF["Trends &\nForecasts"]
        REC["Recommendations"]
    end

    INSIGHTS --> Categories

    style AI fill:#DBEAFE,stroke:#3B82F6
    style INSIGHTS fill:#DCFCE7,stroke:#22C55E

Workflow

1. Nominate Workbooks

Select workbooks for analysis. Only nominated workbooks are included in insight generation.

  • Toggle individual workbooks on/off
  • Bulk nominate/denominate multiple workbooks

2. Generate Insights

Request insight generation for a specific stakeholder view:

Stakeholder Focus
Executive High-level KPIs, strategic trends, board-level metrics
Finance Detailed financial metrics, variances, accounting insights
Operations Efficiency metrics, resource allocation, operational performance
Analyst Data patterns, statistical insights, detailed analysis

3. Review Insights

Each insight includes:

  • Category: Financial health, operational efficiency, risk & compliance, trends, or recommendations
  • Confidence Score: 0.0 to 1.0 indicating analysis confidence
  • Priority: Ranking within the session (1 = highest)
  • Supporting Data: Underlying metrics and data points
  • Methodology: How the insight was derived

4. Ask Questions

Ask natural language questions about:

  • Specific insights: Drill deeper into a particular finding
  • General data: Ask questions about nominated workbook data

Insight Categories

Category Display Name Description
financial_health Financial Health Portfolio-level financial metrics and trends
operational_efficiency Operational Efficiency Process and workflow optimization opportunities
risk_compliance Risk & Compliance Risk exposures and compliance concerns
trends_forecasts Trends & Forecasts Data patterns and forward-looking analysis
recommendations Recommendations Actionable improvement suggestions

Session Management

  • Each insight generation creates a session
  • Sessions can be regenerated up to 3 times for different perspectives
  • Session states: pending -> generating -> completed (or failed)
  • Historical sessions are preserved for comparison

Security

  • Data is processed through SecureAIGateway with PII scrubbing
  • Insights are tenant-isolated
  • Q&A history is per-user and not shared
Page Description
AI Assistant Chat interface, PII handling, limitations

API Reference