Show HN launch page

Bendit is a local DuckDB workbench for messy data files.

I built it for the annoying middle ground between Excel and a full database pipeline: files that are too messy for Excel, but not worth setting up SQL Server, Postgres, Airflow or a full ETL stack.

Download Windows build GitHub Sample CSV

Current state

Import Excel, CSV and Parquet files
Inspect columns and data types
Write DuckDB SQL transformations
Preview results in a grid
Export clean data
Runs locally on Windows
No account required
Downloads through GitHub Releases
Support via email

What it is not

Not a BI dashboard
Not a cloud ETL platform
Not a replacement for a full database IDE
Not finished yet

HN post title

Show HN: Bendit – a local DuckDB workbench for messy Excel/CSV/Parquet files

First comment draft

Hi HN,

I built Bendit because I often receive messy Excel and CSV files where Excel is not enough, but setting up a full database or ETL pipeline feels too heavy.

Bendit is a local desktop workbench built around DuckDB. The idea is to import Excel, CSV or Parquet files, inspect the structure, write SQL transformations, and export clean results again.

It is aimed at the boring but common problem of “I have a messy file and need to turn it into usable data”.

Current state: Windows build, Excel/CSV/Parquet import, DuckDB SQL engine, SQL editor, result grid, export transformed data, GitHub Releases for downloads, GitHub Issues for support, no account required.

I’d love feedback on whether the workflow makes sense, what transformation features are missing, and whether this should stay SQL-first or add a visual pipeline builder later.