Clear Unique Headers
- Requirement
- Required
- Example
- Revenue
- Purpose
- Allows field identification
Upload structured spreadsheet data and ask questions in plain English to compare values, calculate statistics, identify trends, and summarize results.
VizMint should show the fields, filters, calculations, and source rows used to create each answer so users can review how the conclusion was produced.
Supports structured .csv and .xlsx files. Public examples use synthetic data. Review VizMint's current privacy and retention policy before uploading sensitive information.
Use the fixed preview below to switch between the source spreadsheet and a dashboard sample. Open the full demo in a new tab for every KPI and chart.
Ask Questions About CSV
Interactive preview from the sample business dataset
Switch preview
Total Records
48
rows in sample
Total Gross Revenue
1,544,201
sum of Gross Revenue
Total Target Revenue
1,504,652
sum of Target Revenue
Bin width 5,000 · 48 observations
Full dashboard with all KPI details and 6 charts is on the demo page.
Open full demoThe sample contains 48 business records across 4 products and 4 regions. Charts and KPI cards are generated from the same JSON dataset.
Sample totals: $1,439,545 revenue, $631,847 cost, 4,072 orders, and $354 average order value.
Ask a business question and receive a source-grounded answer that exposes its fields, filters, calculation, and supporting rows.
Natural-language analysis is useful only when the answer can be verified. Users need to see the fields, filters, calculations, and source rows used, plus any ambiguity in the question.
The sales sample contains 48 records across 4 products with $1,439,545 in revenue — ready to answer grouped, trend, and comparison questions.
The workflow is designed around traceable, disciplined answers rather than confident-sounding chat.
Users often know the business question they want to answer but do not know which pivot table, formula, or chart to build. A conversational data workflow lets them ask a direct question and receive a source-based answer.
The page must prove more than generic AI chat. It should show a real uploaded file, a natural-language question, the fields used, retrieved rows, calculations, and a transparent answer.
Follow the same dataset from source rows to the generated dashboard preview. The spreadsheet table, KPI cards, and charts all come from one JSON sample, so every visible result can be checked against the source.
The source preview highlights the exact columns and representative rows used in the analysis.
| Date | Quarter | Product | Region | Gross Revenue |
|---|---|---|---|---|
| 2025-07-01 | Q3 2025 | Analytics Pro | North | 27,060 |
| 2025-07-01 | Q3 2025 | Dashboard Plus | South | 25,200 |
| 2025-07-01 | Q3 2025 | Reporting Suite | East | 19,270 |
| 2025-07-01 | Q3 2025 | Implementation | West | 14,430 |
| 2025-08-01 | Q3 2025 | Analytics Pro | South | 29,756 |
The finished example is generated from the sample dataset rather than an unrelated mockup or stock image.
Bin width 5,000 · 48 observations
VizMint works best when every column has one clear heading and each row represents one consistent record.
| Column | Requirement | Example | Purpose |
|---|---|---|---|
| Clear Unique Headers | Required | Revenue | Allows field identification |
| Date | Recommended | 2026-06-01 | Supports trends and period comparisons |
| Numeric Measures | Required for calculations | Revenue, Cost, Quantity | Supports totals, averages, and comparisons |
| Category Fields | Recommended | Product, Region, Team | Supports grouping |
| Stable IDs | Optional | Order ID | Supports distinct counts and row references |
| Text Description | Optional | Enterprise renewal | Provides context but may not be used for numeric calculations |
Rows matching the question or filter.
Columns used in the calculation.
Totals or averages by selected category.
Comparison between two numeric fields.
Change across valid time periods.
Unusually high or low values under a documented method.
Visible note when data is incomplete or ambiguous.
Ability to ask a related question without re-uploading.
| Chart or Output | Required Fields | Business Question |
|---|---|---|
| Supporting Table | Fields and source rows | Which records support the answer? |
| Comparison Bar Chart | Category and numeric measure | How do groups compare? |
| Trend Chart | Date and measure | How is the metric changing? |
| Scatter Plot | Two numeric fields | Is there a visible relationship? |
| Distribution Chart | Numeric field | How are values distributed? |
| Summary Card | Calculated statistic | What is the direct answer? |
Use clear headers, one record per row, consistent dates, and numeric values stored as numbers.
Select the CSV or Excel file using the page's upload control.
For multi-sheet workbooks, choose the sheet containing the relevant data table.
Review the detected fields and map any column VizMint could not identify confidently.
Check dates, numbers, categories, missing values, and record counts before generation.
Create the supported dashboard, report, presentation, chart, or answer.
Review calculations and limitations, then save, share, or export using the options available in the active plan.
| Metric | Formula or Method | Required Fields | Important Rule |
|---|---|---|---|
| Sum | Total of valid numeric values | Numeric field | Exclude invalid and missing values. |
| Average | Sum ÷ valid record count | Numeric field | Do not include missing values as zero. |
| Distinct Count | Count of unique valid IDs | Stable identifier | Formatting variations can create false duplicates. |
| Period Comparison | Current-period metric − previous-period metric | Date and measure | Requires comparable periods. |
| Correlation | Statistical relationship between two numeric fields | Two numeric fields | Does not prove causation and requires enough valid observations. |
| Problem | Recommended Fix |
|---|---|
| Headers are blank or duplicated | Use short unique names. |
| Numbers are stored as text | Convert them to numeric values. |
| Multiple tables exist on one sheet | Keep one rectangular table per worksheet. |
| Summary rows are mixed with detail | Remove totals before analysis. |
| Dates use inconsistent formats | Standardize dates. |
| The question is vague | Clarify the metric, filter, group, or period. |
Available options depend on the live VizMint plan. The final page should list only verified capabilities, such as saving the analysis, downloading chart images, exporting PDF or PowerPoint files, creating a shareable link, refreshing with updated data, or removing branding on eligible plans.
Do not publish a plan comparison until product and billing owners confirm every feature.
VizMint analyzes exports; it does not replace the accounting, CRM, help-desk, HR, warehouse, advertising, or project-management system that produced them.
Upload a structured CSV or Excel file, review the detected fields, and generate the workflow-specific output described on this page.