What Are Response Visualizations?

Response visualizations turn structured data in the AI’s answer into an interactive table or chart. If the assistant’s reply includes a markdown table (for example, monthly sales, product comparison, or survey results), you can ask to “show as a table” or “display as a bar chart” and FAQ Ally will parse that data and render it in the format you chose. You can then switch between table and different chart types from the same data—no need to re-ask the question.

How to Ask for a Visualization

Include a phrase in your message that describes the format you want. The system recognizes the following (case doesn’t matter):

  • Table — e.g. “show as a table”, “display as table”, “in table format”
  • Bar chart — e.g. “as a bar chart”, “show me a bar chart”, “display as bar chart”
  • Line chart — e.g. “as a line chart”, “line chart”, “display as line chart”
  • Pie chart — e.g. “as a pie chart”, “pie chart”, “show as pie chart”
  • Scatter plot — e.g. “as a scatter plot”, “scatter plot”, “display as scatter”
  • Area chart — e.g. “as an area chart”, “area chart”, “display as area chart”

Example: “Compare our top five products by revenue and show as a bar chart.” If the AI’s reply contains a table (in markdown), that table will be rendered as a bar chart. You can also ask for a table first, then use the format switcher to view the same data as a line or pie chart.

Table View

When the response is shown as a table:

  • Sortable columns — Click a column header to sort by that column (ascending or descending). Useful for ranking, dates, or numbers.
  • Expand to see more — Long tables are capped initially; use “Show more” to expand and see all rows.
  • Consistent layout — Headers and cells align so data is easy to scan and copy.

Chart Types

📊 Bar chart

Categories on one axis and values on the other. Good for comparisons (e.g. products, regions, categories). On smaller screens, bars are shown horizontally for readability.

📈 Line chart

Values over categories or time. When the first column looks like dates (e.g. “Jan 2024”, “2024-01-15”), the x-axis uses a time scale with formatted date labels so trends over time are clear.

🥧 Pie chart

Single series as proportions of a whole. Best when you have one set of values (e.g. market share, budget breakdown).

🔵 Scatter plot

Two numeric columns: first data series as x, second as y. Useful for correlations or two-metric comparisons.

📉 Area chart

Like a line chart with the area under the line filled. Also uses a time scale when the first column looks like dates. Good for cumulative or volume-over-time views.

Switching Format Without Re-asking

Once a response has been visualized, a format dropdown (or format bar) appears above the table or chart. You can switch between Table, Bar, Line, Pie, Scatter, and Area without sending a new message. The same underlying data is re-rendered in the chosen format—ideal for exploring the same answer in different ways (e.g. table for exact numbers, line chart for trend).

Date Detection and Time Scale

For line and area charts, if the first column of the data (labels) looks like dates (e.g. “2024-01-15”, “Jan 2024”, “Q1 2024”, or “1/15/2024”), FAQ Ally uses a time scale on the x-axis. The system checks a sample of the labels and, when most look like dates, switches to the time scale. Axis labels are formatted as dates (e.g. “Jan 2024”, “MMM yyyy”) so time-based trends are easy to read. Bar, pie, and scatter charts do not use a time axis.

Full-Screen and Mobile

In the on-premise (in-app) chat, you can open the chart in a full-screen overlay for a clearer view. On narrow screens (e.g. phones), bar charts in the on-premise chat automatically use a horizontal layout and touch-friendly controls so the visualization stays usable. The embeddable widget offers the same table and chart types with a format switcher; chart rendering is responsive to the widget size.

Data Quality and Sources

When a visualization is built from the AI’s answer, the UI can show how many sources were used (e.g. “Based on 3 sources”). If the system detects low confidence for the answer, a warning may appear so you know to double-check the data. Empty or missing data is handled gracefully with a clear “No chart data to display” state.

When to Use Visualizations

  • Comparisons — “Compare X and Y, show as bar chart.”
  • Trends over time — “Sales by month as a line chart.” (Dates in the first column get a time axis.)
  • Proportions — “Break down budget by category as a pie chart.”
  • Two metrics — “Plot cost vs. revenue as a scatter plot.”
  • Exact numbers first — “Show as a table” then switch to a chart to see the same data visually.

Visualizations work in on-premise chat, the embeddable widget, and via the Chat API (the API returns the same visualization payload with tableData and chartData for you to render). The AI’s response must include a markdown table for the system to parse and display. For best results, ask questions that naturally lead to tabular or numeric answers (e.g. “What were our top 5 products by revenue last quarter?” or “List monthly sign-ups for the past six months”).