Spreadsheet — work directly with your data
Edit cells, join datasets, build pivots, and calculate new columns in a familiar grid. Commit your changes to capture them as repeatable steps in your analysis.
From raw data to results you can retrace.
Explore like a spreadsheet. Explain like a notebook. Present like a dashboard. Move between views without losing your place. Or let AI plan your analysis — then review and refine each step before it runs. Bring your own cloud key, keep AI local, or work without it.
Free beta · Fully usable with the AI switched off
Notebook — see how you got there
Follow your analysis in readable code, add your own calculations, and explain your thinking with notes. Keep the method alongside the results, ready to revisit and rerun.
Dashboard — make your results clear
Turn your analysis into charts, filters, and notes. Explore the patterns, bring your findings together, and export a report others can use.
The pipeline — build, review, repeat
See how your analysis fits together, step by step. Add operations yourself or review a plan proposed by AI, then edit, undo, or replay your work on the next dataset.
Review before you run
The agent proposes a visual plan with a preview of each step's results. Refine the steps, check the changes, and approve when you're ready. Your data changes only when you apply the plan.
Build once. Run again.
Save your analysis and replay it on next month's file. The same steps run in the same order, giving you a repeatable process you can inspect and refine.
Your machine does the work
All data processing happens on your machine. Cloud AI receives column names, profile statistics, and the analysis plan — never raw rows. With local AI, inference stays on your machine too.
| Mode | What it means |
|---|---|
| Cloud AI | Connect your preferred model with your own API key. Pay your provider directly. |
| Local AI | Run a model on your machine. Keep inference on-device, with no API fees. |
| No AI | Work directly with your data. The spreadsheet, notebook, dashboard, and pipeline are all yours to use. |
Switch whenever you like. Because a prompt carries your schema and profile statistics rather than your rows, the cost of a step tracks how wide your data is — not how many rows are in it.
Full access. Run it with no AI, with a local model, or with your own cloud key — any model spend is yours, billed by your provider.
Paid plans arrive once the beta wraps — shaped by what beta users actually need.
Good analysis should be easy to follow — and build on.
DataGadt is built for analysts and data scientists who want to keep their data, methods, and results together. It brings the everyday work of cleaning, exploring, and presenting data into one workspace on your machine.
At its centre is a pipeline that makes the process visible. Work directly with your data or let AI propose the next steps — you stay in control of what runs. The aim is simple: analysis you can understand today, repeat next month, and hand over with a clear account of how it was built.
Download the beta
Beta access is by invite — log in if you have an account, or request one below.
Windows and macOS. Neither build is code-signed yet, so the first launch takes one click past the operating system's warning — the download page walks you through it.