Insights Studio - User Guide

How to create dashboards and charts from your agent data (step-by-step)

What is Insights Studio?

Insights Studio helps you extract meaningful business insights from your custom agent data — without spreadsheets or complex analytics tools.

In a few clicks, you can create dashboards, explore your dataset, and build charts to track patterns such as volume, trends, category splits, and performance over time.

Who is this for?

•  Business owners and managers who want quick answers from agent data

•  Operations teams who track daily activity and outcomes

•  Any non-technical user who needs charts and dashboards in minutes

What you can do?

•  Create dashboards from any agent dataset

•  View your raw data in Data Explorer

•  Build charts (Bar, Line, Pie, Donut, Area, Scatter)

•  Group by time (Hour, Day, Week, Month)

•  Choose metrics (Count, Sum, Average, Min/Max, etc.)

•  Add filters to focus on specific segments

In this guide, we will walk through the full flow:

1. Open Insights Studio and pick an agent dataset.

2. Create a dashboard.

3. Explore the raw data in Data Explorer.

4. Create charts with the chart builder.

5. Use filters and time grouping to answer business questions.

1. Open Insights Studio and select an agent

From the left sidebar, open Insights Studio. You will see a list of available agents (datasets). Each agent represents one type of data your organization is collecting.

Insights Studio home page – Available Agents and the “Create dashboard” action.

To start:

• Step 1: Find the agent you want (example: “Visitor Entry Extraction”).

• Step 2: Click “Create dashboard”.

2. Create a dashboard

After you click “Create dashboard”, a popup will open.

• Enter a dashboard name (example: Visitor Entry Analytics).

• Click “Create dashboard”.

Now you will land on the dashboard editor screen.

Dashboard editor – Time filter pill, Charts/Data Explorer tabs, and “Add Chart”.

On this page you will mainly use these controls:

• Time filter pill (example: “All Time”) – quickly limit your analysis to a time range.

• Tabs: Charts and Data Explorer – switch between visual charts and the raw table data.

• Refresh Data – reload the latest dataset from the agent.

• Add Chart – create a new visualization for this dashboard.

3. Filter data by time

Use the time filter pill to focus on a specific period such as Today, Last 7 days, or Last 30 days.

Time filter pill – choose All Time, Today, Yesterday, Last 7/14/30 days.

Tip: If your dashboard looks empty, set the time filter back to “All Time” to confirm that data exists.

4. Explore your data in Data Explorer

Before building charts, it is helpful to first understand what fields (columns) are available in the dataset. Use the Data Explorer tab to view the raw table.

Data Explorer – view your agent dataset as a table with all available columns.

What you can do here:

• Scroll to see rows (records) and columns (fields).

• Confirm which column contains time/date (useful for trends).

• Identify the business fields you want to analyze (example: project_name, purpose_of_visit, store, coupon, etc.).

5. Create your first chart

Click the “Add Chart” button to open the chart builder. This is where you configure how your visualization should look and what it should measure.

Create New Chart – chart name, chart type, dimensions, time grouping, metrics, and optional filters.

The chart builder works in this order:

• Step 1: Give the chart a name.

• Step 2: Choose a chart type (bar/line/pie/etc.).

• Step 3: Select Dimensions (what goes on the X-axis, and optional breakdown).

• Step 4: Optionally group by time (hour/day/week/month).

• Step 5: Choose a Metric (count or aggregation like sum/average).

• Step 6: Optionally add Filters to narrow down the data.

• Step 7: Click “Create Chart” (or “Update Chart” if editing).

5.1 Choose a chart type

Pick a chart type based on what you want to understand:

• Bar: compare categories (example: visits by project, expenses by category).

• Line / Area: show trends over time (example: daily orders).

• Pie / Donut: show share or distribution (example: purpose of visit).

• Scatter: compare two numerical fields (example: discount percent vs effective price).

Chart Type options – Bar, Line, Pie, Donut, Area, Scatter.

5.2 Set Dimensions (X-axis and Breakdown)

Dimensions decide what your chart is grouped by.

• X-Axis (Primary Dimension): the main field you want to analyze.

• Breakdown (Optional): splits each X-axis item into multiple series (useful for comparisons).

Dimensions – select the X-axis field and an optional Breakdown field.

Example ideas:

• Orders by Store: X-axis = store, Metric = count (or sum of revenue).

• Purpose of visit by Project: X-axis = project_name, Breakdown = purpose_of_visit.

• Daily visits: X-axis = date/created_at, Time Dimension = day, Metric = count.

Column selector – choose any available dataset column as a dimension or metric input.

5.3 Use Time Dimension for trends

Time Dimension is optional. Use it when you want to group results by time (hour/day/week/month).

• None: no time grouping (best for category comparisons).

• Hour: results grouped into hours (9 AM, 10 AM, etc.).

• Day: results grouped day-by-day.

• Week: results grouped week-by-week.

• Month: results grouped month-by-month.

Note: For time grouping to work, your selected X-axis should be a date/time field (for example: created_at, date, entry_time).

5.4 Choose a Metric (what to calculate)

Metrics decide what number the chart should show for each group.

• Row Count (COUNT *): counts how many records exist in each group. This is the most common choice.

• Column Aggregation: calculates a value using a selected column and an aggregation function (sum, average, minimum, maximum, etc.).

Metric – switch between Row Count and Column Aggregation, and pick an aggregation function.

When to use which:

• Use Row Count when you want “how many” (how many visits, how many orders, how many records).

• Use Column Aggregation when you want “how much” or “how big” (total revenue, average discount, max bill amount).

5.5 Add Filters (optional)

Filters narrow down your data so the chart answers a more specific question.

• Example: show only one project, only one store, or only records with discount_percent > 0.

• You can add multiple filters. All filters apply together.

Filters – add one or more conditions to narrow down the chart data.

Tip: Start without filters. Once the basic chart looks right, add filters step-by-step.

6. View and use your dashboard

After clicking “Create Chart”, close the chart builder and return to the dashboard. Your chart will appear in the Charts tab.

• Use the time filter pill at the top to quickly change the time window for all charts.

• Click “Refresh Data” to pull the latest data if new records were added.

• Create multiple charts in the same dashboard to answer different questions.

Good dashboard habits:

• Keep chart names clear and action-oriented (example: “Daily visitor count”, “Top projects by visits”).

• Prefer 5–10 charts per dashboard. Too many charts make it harder to read.

• If a chart has too many categories, add a filter or switch to a time trend.

7. Ready-to-use chart recipes (examples)

Below are common charts that business users typically create. You can adapt them to any dataset.

7.1 Visitor Entry (site / project operations)

• Daily visitor count: Line, X-axis = date/created_at, Time Dimension = Day, Metric = Row Count.

• Visitors by purpose: Pie/Donut, X-axis = purpose_of_visit, Metric = Row Count.

• Visitors by project: Bar, X-axis = project_name, Metric = Row Count.

• Peak visiting hours: Bar, X-axis = entry_time, Time Dimension = Hour, Metric = Row Count.

7.2 Sales Tracker (revenue and performance)

• Total revenue by store: Bar, X-axis = store, Metric = Column Aggregation (sum of effective_price or revenue).

• Orders trend: Line, X-axis = created_at, Time Dimension = Day, Metric = Row Count.

• Discount impact: Scatter, X-axis = discount_percent, Metric = (optional) aggregation of effective_price.

7.3 Expense Tracker (cost control)

• Expenses by category: Bar, X-axis = category, Metric = Column Aggregation (sum of amount).

• Monthly expenses trend: Area, X-axis = date, Time Dimension = Month, Metric = Column Aggregation (sum of amount).

8. Troubleshooting

If something does not look right, try these quick checks:

• Chart is empty: set time filter to All Time, and click Refresh Data.

• Trend chart looks wrong: confirm you selected a date/time field on X-axis and chose a Time Dimension.

• Too many bars: add a filter (example: one store) or switch to a time trend.

• Number looks too high/low: verify Metric type (Row Count vs Column Aggregation) and the aggregation function.

• Data changed recently: use Refresh Data to reload the latest dataset.

9. Glossary (simple definitions)

• Agent / Dataset: the source of data (example: Visitor Entry Extraction).

• Dashboard: a collection of charts created from one dataset.

• Chart: one visualization (bar/line/pie, etc.).

• Dimension: a field used to group data (X-axis and optional breakdown).

• Metric: the value calculated for each group (count, sum, average, etc.).

• Aggregation: a way to combine many rows into one number (sum/average/min/max).

• Filter: a rule to include or exclude records in a chart.
  
  

  

  Dashboard – see how your charts populate.

Need help? Share your dashboard name and the question you want to answer, and we can suggest the best chart setup.