Data Analysis Coordinator

What it solves

Table analysis often breaks when column names do not define business metrics. A few leading rows may hide duplicate rows, merged cells, inconsistent time granularity, or regional rollups. This skill treats those as process problems: it separates requirement clarification, parsing, quality checks, aggregation, charting, and verification before claims are written.

How it works

• Preflight and scoping: checks .xlsx, .xls, .csv, sheets, encoding, merged cells, and readability; extracts metrics, dimensions, time range, filters, and audience.
• Reliable calculation: uses full data or verifiable aggregations for groups, time series, Top/Bottom, YoY/MoM, and total reconciliation; avoids extrapolating from a 20-row sample.
• Charts and reports: produces Markdown and Word outputs with titles, units, time, source, and notes; falls back to tables and text when charts fail.
• Limits disclosure: missing values, anomalies, filters, cleaning, and metric scope are recorded; hypotheses are not presented as proven causality.

Where it fits

Best for tabular data with a decision-oriented question. Files above 100 MB or 100,000 rows should be chunked or aggregated in a database; fewer than 10 rows, heavy key-field loss, or weak time coverage require weak descriptive conclusions only. Non-tabular files, pure cleaning tasks, or sensitive external data transfers need different tools or explicit data-safety checks.