Chart Advisor

Problem

When you have sales, ops, or finance data but are unsure whether to use a bar chart, line chart, or table, Chart Advisor narrows the task into confirmable choices. It does not replace OCR or deep statistical modeling; it focuses on the data-to-chart/table delivery path: identify the reader role, infer the goal, and present 2-3 practical directions instead of overwhelming the user.

How It Works

The core flow is: role → goal → template → preprocessing → charting → validation → delivery.

• Role calibration: infers whether the user is rushed, partially prepared, or experienced, then switches between simplify, standard, or empower modes.
• Goal anchoring: after receiving data, asks in plain language what you want to know, such as ranking, trend, period comparison, or anomalies.
• Template and table: selects a column structure from T1-T6, such as ranking, comparison, time series, or KPI overview, and confirms fields before processing.
• Deliverable: generates a multi-sheet .xlsx workbook with raw data, summary, embedded chart, and optional splits; the script applies number formats, totals, embedded charts, and a Ctrl+T table.
• Validation: checks file existence, embedded charts, data integrity, formatting, required sheets, and cross-dimension consistency before delivery.

Boundaries

It is useful for chart selection, Excel table standards, color, and title guidance. It is not intended for screenshot parsing or complex modeling such as regression, clustering, or forecasting; aggregate large category sets first. Color, title, and chart type can still be adjusted after delivery.