demo.mp4

Savvy

Local tool for working with SPSS survey files. Load a .sav file, explore columns and value labels, configure what to export — all in a browser UI that runs entirely on your machine. No data leaves your machine.

Built on pyreadstat for SPSS file parsing.

Usage

Download a release (recommended for end users)

1. Install uv
2. Download the latest savvy-vX.X.X.tar.gz from Releases
3. Extract and run:

tar -xzf savvy-v0.1.0.tar.gz
cd savvy
uv run uvicorn backend.main:app --port 8000

4. Open http://localhost:8000

From source (developers)

git clone https://github.com/cedanl/savvy
cd savvy
make install
make dev

Open http://localhost:5173. If running via devcontainer-cli (plain Docker), run make url to get the correct address.

What it does

After loading a .sav file:

• See all columns with their full question labels, detected type, and sample values (chips update live as you change the export mode)
• Search and filter columns by name or label
• Toggle columns on/off with select all / deselect all
• For columns with value labels: choose per column whether to export raw codes (Numbers), readable labels (Text answers), or both
• For columns with both a question label and a coded name: pick which to use as the CSV header
• Export to CSV

API

Endpoint	Description
POST /api/file	Upload a .sav file, returns column metadata
POST /api/export	Upload a .sav file + JSON manifest, returns CSV

Development

Open in a devcontainer for a pre-configured Python 3.13 + Node 22 environment.

make test       # run all tests (backend + frontend)
make backend    # backend only, port 8000
make frontend   # frontend only, port 5173

Tests: 19 backend (pytest) + 68 frontend (vitest).

License

EUPL v1.2