> ## Documentation Index > Fetch the complete documentation index at: https://docs.replit.com/llms.txt > Use this file to discover all available pages before exploring further. # Create a dashboard from data > Turn a public CSV, spreadsheet, database, warehouse connector, or API into a dashboard that answers a clear question. export const AiPrompt = ({children}) => { return
{children}
; }; A dashboard helps someone make a decision from data. You will build a **Global Progress Explorer** from a public CSV: country-level data with year, population, continent, life expectancy, and GDP per capita. ## What you'll achieve By the end of this guide, you will have: * A Global Progress Explorer dashboard built from a realistic public CSV. * Summary metrics for countries, population, life expectancy, and GDP per capita. * Charts and filters that show how progress changed over time. * A way to verify that the dashboard numbers match the source data. * A shareable or publishable dashboard for stakeholders. ## What you'll learn You will learn how to: * Start with a decision, not a chart type. * Ask Agent to build a dashboard from a public CSV. * Check calculations against the underlying data. * Add filters and an insights summary. * Decide when a dashboard needs login, refresh, exports, or connected data. ## Who this is for Use this guide if you have customer, sales, product, operations, research, or public data and need a dashboard people can understand. This is useful for analysts, operators, founders, PMs, educators, and teams who need to turn raw data into decisions. ## Define the decision Before building, write down the decision the dashboard should support. For the global progress dataset: * **Question**: Where did life expectancy and GDP per capita improve over time? * **Audience**: education, research, strategy, or nonprofit teams. * **Useful metrics**: country count, year range, latest median life expectancy, total population represented, and median GDP per capita. * **Useful filters**: continent, country, and year range. If you cannot state the decision, the dashboard will become a collection of charts instead of a useful tool. ## Build from a public CSV Use a positive public dataset with meaningful fields and enough rows to explore. Dataset: [Gapminder global progress CSV](https://raw.githubusercontent.com/plotly/datasets/master/gapminderDataFiveYear.csv) Ask Agent to build the dashboard from that CSV: Build a Global Progress Explorer dashboard from this public CSV:
[https://raw.githubusercontent.com/plotly/datasets/master/gapminderDataFiveYear.csv](https://raw.githubusercontent.com/plotly/datasets/master/gapminderDataFiveYear.csv)
The dashboard should help an education or research team understand how life expectancy, population, and GDP per capita changed across countries and continents from 1952 to 2007.
Include summary cards for total countries, year range, latest median life expectancy, total population represented, and latest median GDP per capita.
Add a line chart for life expectancy over time by continent.
Add a bubble chart where GDP per capita is on the x-axis, life expectancy is on the y-axis, bubble size represents population, and color represents continent.
Add a bar chart showing the countries with the largest life expectancy gains between their first and latest year in the dataset.
Add filters for continent, country, and year range.
Include a short insights summary that explains the biggest patterns and which regions or countries improved most.
Show how key metrics are calculated so I can verify the numbers.
Make the dashboard polished, responsive, and easy to read on desktop and mobile. ## Check your work Open Preview and verify the dashboard before sharing it. Check that: * Total countries matches the number of unique countries in the CSV. * The year range shows 1952 to 2007. * Latest median life expectancy uses the latest year in the dataset. * Population totals use the filtered countries and years. * Filters update the summary cards and charts. * The bubble chart uses GDP per capita, life expectancy, population, and continent correctly. * The insights summary matches what the charts show. * The dashboard works on a phone-sized screen. If a number looks different than expected, ask Agent to show the calculation. The latest median life expectancy looks different than I expected.
Show me how you calculated it from the CSV.
Check that you used only rows from the latest year in the filtered data.
Update the calculation if needed and tell me how to verify it. ## Add filters and exports After the first dashboard works, ask Agent for the interaction that helps the audience explore the data. Add an exploration panel for the global progress dashboard.
Let me filter by continent, country, and year range.
Add a button to export the filtered chart data to CSV.
Keep the summary cards and insights summary in sync with the filters. Dashboards can also support PDF export, chart CSV export, refresh, auto-refresh, light/dark mode, and analysis summaries. See [Data Visualization](/references/artifact-types/data-apps) for the concept overview. ## Use connected data when needed Use a public CSV when you are learning or prototyping. Use a connected data source when the dashboard should stay up to date. You can connect dashboards to: * A Replit app database. * A warehouse connector like BigQuery, Databricks, or Snowflake. * Analytics tools like Segment, Amplitude, or Hex. * External APIs. For warehouse data, ask Agent to explain the fields and queries it uses. Create a growth dashboard using our warehouse connector.
Identify the customer, product usage, billing, and retention fields needed.
Build charts for active customers over time, account growth by segment, and expansion revenue by plan.
Explain the fields and calculations you used so I can verify the dashboard. For setup and availability details, see [Data Connectors](/connectors/warehouses/overview). ## Share the dashboard Share the dashboard when the numbers are trustworthy. If the dashboard uses private customer or business data, add access control before sharing widely. For public or sample data, you can publish and share the URL directly. Ask reviewers: > Review the global progress dashboard and tell me which region or country story you would highlight first, and why. ## You're done when * The dashboard loads the public CSV. * Summary cards and charts answer the progress question. * You verified at least one key metric manually. * Filters update the dashboard correctly. * The insights summary matches the data. * You know whether the dashboard should stay public, require login, or connect to live data. ## Next steps Put the dashboard behind authentication so only the right people see it. Save data so it persists and refreshes between sessions. Pull in another data source or push insights to existing tools. Connect a live warehouse so the dashboard reflects up-to-date data. Ask Agent to create a stakeholder presentation from the dashboard's findings. ## Need further help? * **Numbers look different than expected:** ask Agent to show the query, formula, or transformation. * **The CSV is unavailable:** check the URL, network access, and CSV headers. * **Filters do not update charts:** ask Agent to connect filters to both summary cards and chart data. * **The dashboard is slow:** ask Agent to summarize data, add filters, or reduce rows loaded on startup. * **The dashboard uses private data:** add login before sharing broadly. ## Related * [Data Visualization](/references/artifact-types/data-apps) * [Data Connectors](/connectors/warehouses/overview) * [Add login](/build/add-login) * [Add integrations](/build/add-integrations)