The workflow
How Corgage works: from file upload to finished report
Corgage replaces the slowest parts of an analysis request: cleaning the file, understanding the columns, and writing up what it means. Here is what happens between the upload and the answer.
1. Upload anything you already export
CSV, Excel workbooks and JSON all work, and you can select several files at once. Corgage reads them in the browser, so your rows stay on your machine while the analysis runs.
2. Scattered files become one dataset
Titles, blank rows and notes above the real header are detected and skipped. When two files describe the same thing with different column names, Corgage matches them and keeps a source column so you can always trace a row back.
3. Every column gets profiled
Corgage detects the type of each column, then reports distributions, medians, ranges, missing values, duplicates and outliers. Numeric columns are correlated against each other so relationships surface without you writing a formula.
4. Ask the analyst
The chat assistant is grounded in your dataset's real statistics, not a guess. Ask why revenue dipped, request a written summary for a board update, or have it draft the commentary that goes with a table.
5. Keep it running daily
Each upload is fingerprinted and compared to the previous snapshot. Rows added, columns changed and metric shifts appear as a timeline, and you can ask for a briefing on what moved since yesterday.
Try it on your own export
Bring the messiest file you have. Plans start at $49 a month after a 3-day free trial. You can also run a free CSV analysis first, read our data analysis guides or check the FAQ.