1. Interpreted business question
PlannedQuestion intent parsing and business term normalization.
AI Intelligence Investigation Workspace
Central differentiator: natural-language business investigation that keeps evidence, assumptions, and limitations explicit.
Investigation Vision
Question to evidence workflow for executive and analyst use
Business question
Why did revenue decline in Germany?
Investigation
Structured analysis across commercial, operational, and macroeconomic dimensions.
Result
Findings, supporting evidence, confidence limits, and suggested follow-up questions.
Question Workspace
Interface scaffold only. No AI execution or simulated analytical conclusions are generated in current build.
AI provider and SQL execution backends are not connected yet. This page currently defines interaction and audit structure.
Investigation Scope
Planned analysis domains for each business question
Revenue development
Product contribution
Customer contribution
Sales volume evolution
Pricing and margin movement
Relevant external economic indicators
Investigation Lifecycle
End-to-end explainable investigation flow
2. Investigation plan
PlannedStepwise plan with hypothesis and validation points.
3. Relevant tables and definitions
AvailableMetadata can already map SAP-inspired table structures.
4. Executed SQL queries
Not ConnectedQuery execution log UI is prepared but backend execution is not connected.
5. Results and evidence
Not ConnectedResult cards and evidence traces will be populated by backend outputs.
6. Charts and breakdowns
PlannedMulti-granularity visual analysis blocks are designed but not wired.
7. Conclusions and confidence
PlannedNarrative conclusion with confidence scoring is reserved for backend logic.
8. Suggested follow-up questions
PlannedFollow-up prompt generation is scaffolded for future AI integration.
Planned SQL Evidence Log
Query lineage and reproducibility
SQL execution records will appear here with query text, execution time, row counts, and table lineage once backend integration is available.
Result Framework
How findings, correlations, and hypotheses will be distinguished
Verified findings
Not ConnectedStatements supported by executed queries and traceable evidence.
Correlations
PlannedObserved patterns that may explain performance but do not prove causality.
Hypotheses
PlannedCandidate explanations to test in follow-up analysis.
Confidence and limitations
PlannedExpected confidence scoring and explicit data limitations in final narrative.