Experiments with Rotating Nested Circles
By Curran Kelleher
VizHub AI Studio is in Early Access
Go beyond dashboards. Create custom interactive visualizations with AI.
Bring your data, describe what you want to explore, and compare working AI-generated alternatives. Refine the result, publish it, or export editable React and D3 code you own.
Explore a real interactive example
A public VizHub visualization, not a claim that this example was generated by AI.
From dataset to interactive experience
See how an idea becomes a visualization
A real AI Studio workflow with sample data: start with a question, review the alternatives, and keep the result. No fixed chart templates required.
01Start with your data
Add a dataset and explain the question you want to explore. VizHub keeps the data and visualization in one project.
02Compare AI alternatives
Review working proposals before choosing what to keep. You stay in control of which changes are applied.
03Refine and share the result
Keep improving the interactive visualization, publish it, or export the editable source to use elsewhere.
Product screenshots from a local demonstration using public sample economic data. These images illustrate the workflow; they are not live previews.
Featured visualizations
Explore interactive projects created and shared on VizHub. These examples are not necessarily generated using AI Studio.
Meet the founder
Built by people who build visualizations
Curran Kelleher is the founder of VizHub, a data visualization engineer, consultant, and educator. He previously worked at Stamen Design and McKinsey & Company and teaches data visualization at Worcester Polytechnic Institute.
VizHub brings that experience into a workspace for designing interactive visualizations, collaborating on ideas, and retaining control over the code.
Explore professional services →Why build with VizHub?
Standard chart builders focus on predefined layouts. VizHub combines data-aware AI, React and D3, reviewable alternatives, and exportable source code for custom interactive work.
Understands your data
CSV and JSON datasets are profiled automatically, so AI edits are grounded in your actual schema and values.
Code you can take with you
Export editable visualization source files as a ZIP and keep building outside VizHub.
Compare AI alternatives
Generate alternatives with multiple AI models, review the results, and approve the version you want to keep.
React + D3 flexibility
Start with React and D3, then edit the code to build interactions beyond standard dashboard charts.
Built for teams that work with data
Explore four dataset-backed example projects, from county health indicators and ecommerce events to storm reports and economic research. Each brief includes source data, interactions, and a review checklist.
Research organizations
Explore county health, poverty, education, and income through linked public-data views.
Example: County Health and Opportunity Explorer
Explore dataset-backed projectVisualization consultancies
From one synthetic ecommerce brief, explore three distinct interactive directions.
Example: One Dataset, Three Designs
Explore dataset-backed projectNewsrooms
Explore 2024 NOAA storm reports through regional comparisons and timelines.
Example: A Year of Extreme Weather
Explore dataset-backed projectFinancial research publishers
Explore inflation, unemployment, and interest rates with linked economic charts.
Example: The Economic Pulse
Explore dataset-backed projectBring your own agent
Works with the AI tools you already use
Use VizHub's built-in AI, or connect external coding agents through MCP. People and agents work with the same project, data, and revision history rather than separate one-off outputs.
VizHub's built-in AI
Generate and refine visualizations without leaving the workspace.
External agents
Connect compatible coding agents through MCP to work with your VizHub project.
One shared project
Authorized agents can work with the same project data and visualizations you review in the browser.
Data and access
Understand how VizHub handles your work
Review visibility, AI processing, and data-use guidance before you upload a dataset or publish a visualization.
Choose who can see your work
Published public work can be discovered. Unlisted standalone work is link-accessible; project permissions may further restrict access. Free owners can unpublish to unlisted. Creating private personal work requires Pro.
Know when AI receives your data
When you request an AI edit, the prompt and relevant project content are sent through OpenRouter to third-party model providers.
Check data suitability first
Do not upload regulated or highly sensitive information without a separate written agreement. Review the data and privacy guidance before starting.
Create something interactive. Keep the code.
Turn your data into a custom visualization, refine it with AI, and publish or export the source files to keep building in your own project.
Start building freeWant to understand the process first? Read the quick-start guide.
Need a custom implementation instead? Work directly with a visualization specialist.
Build better visualizations together
Join the VizHub community to share your work, exchange ideas, get help, and influence what we build next.