Smart Data Visualization

Problem

Data often sits in Excel, CSV, JSON, screenshots, or report images. Cleaning fields, choosing chart types, and rebuilding reports manually is repetitive and hard to keep consistent. This skill targets the workflow from raw data to a shareable chart report, reducing manual parsing, chart selection, and layout work.

How It Works

• Multi-source parsing: supports .xlsx, .csv, .json, .txt, and table extraction from screenshots or images.
• Automatic analysis: detects numeric, date, categorical, and boolean columns; checks missing values, duplicates, and outliers; computes means, standard deviations, correlations, and trends.
• Chart recommendation and rendering: selects line, bar, scatter, heatmap, Sankey, or other charts based on data features and produces interactive charts with zoom, filtering, and export.
• Report output: generates HTML, Excel, PDF, or image exports for demos, secondary analysis, and formal documentation.

Boundaries

Best for tabular or semi-structured data involving trends, comparisons, distributions, and proportions. Ambiguous screenshots and complex business rules may still need human review. Anomaly detection and forecasting are decision aids, not substitutes for validated thresholds and data quality checks.