Bedevere
A tabular data visualizer and DuckDB SQL editor for the browser. Use it as a standalone web app, or embed it in your own page
— as an @kolistat/bedevere-wise npm component, or a single /embed iframe with no build step.
What it does
- Open CSV, Parquet, JSON, and Arrow files directly in the browser — no upload, no server round-trip.
- Write DuckDB SQL against them with a real editor: syntax highlighting, completion, and instant results.
-
Pivot, filter, and chart the result table without leaving the page — or plot straight from SQL with the
VISUALIZEgrammar (ggsql), which now aims to follow the grammar of DuckDB’s ggplot extension. -
Embed the same surface in your own product — the
@kolistat/bedevere-wisenpm component, or a one-line/embediframe you prefill with URL parameters (no build step, no backend).
Where it fits
Bedevere is for any analyst who needs to look at a table, run a query, and ship the answer — fast. It runs on the same in-browser DuckDB engine that powers the rest of KoliStat, tuned for interactive exploration: open a file, write SQL, visualize, done.
Bedevere Desktop
Prefer a native app? Bedevere Desktop runs real DuckDB on your own machine — full speed, bigger datasets, and files opened in place — in about 75 MB, with no browser sandbox. It’s local-first: your data never leaves your device.
Free during the beta. Windows 10/11 (x64); the installer isn’t code-signed yet, so Windows may warn you — choose More info → Run anyway.
Your use is governed by the License · Privacy · Third-party notices.
Links
- Live app: bedeverewise.app
- npm package: @kolistat/bedevere-wise
- Source: github.com/KoliStat/bedevere-wise
Need a feature?
Bedevere is shaped by the people who use it. If your team needs something it doesn’t do yet — a format, an export, an integration — reach out. Feature requests and sponsored work are both welcome.