insight_plan

A phased data insight workflow decomposes complex analytical questions into 2-4 structured milestones (L1-L4) to guide root-cause analysis. This approach sequences multi-step investigations, ensuring each phase builds upon previous findings for structured decision support.

To plan a macro-level data investigation, generate a MacroPlan JSON that splits your user question into 2-4 analysis phases. The plan defines phase sequencing and focus areas, providing structured guidance for multi-step anomaly investigations and KPI storytelling.

No database access is required during data insight plan generation. The planning process relies entirely on domain knowledge and provided references to decompose questions into phased milestones, outputting a structured JSON MacroPlan without querying any live data sources.

Yes, structured phase decomposition supports anomaly investigations and KPI storytelling. By generating a MacroPlan with 2-4 phased analyses, you can sequence root-cause investigations and structure downstream steps for macro-level data inquiries and insights projects.

A MacroPlan for data-insight tasks is limited to planning phase decomposition and does not execute the actual data analysis. It splits questions into 2-4 phases (L1-L4) but requires downstream execution tools to query data and validate the root-cause analysis milestones.