RowcallEarly beta

A Python workspace for exploring, transforming, and building with data.

Rowcall combines interactive exploration with an explicit graph of Python steps. See how your analysis fits together, inspect results as you go, and keep your work in ordinary Python files.

Try the beta View on GitHub ↗
An analysis, step by step
≡Orders✓
pl.read_csv("orders.csv")
orders ●
≡Customers✓
pl.read_csv("customers.csv")
customers ●
ƒJoin datasets✓
orders.join(customers,
on="customer_id")
● joined ●
ƒRevenue by segment✓
joined.group_by("segment")
.agg(pl.col("revenue").sum())
● segment_revenue
Result preview

Revenue by segment

Illustrative data · 3 rows
Consumer$48,200
Business$34,050
Wholesale$19,900

A quick investigation rarely stays simple.

You bring in another source, join datasets, try a different transformation, and build a model. Rowcall keeps those steps connected in a graph, so you can see how a result was produced and build on the work.

In a notebook, the order you run cells can differ from the order you read them. In Rowcall, connections make dependencies visible and determine which steps run.

Features

Traceable, reproducible analysis

See which steps produce each result. Rowcall runs the required steps in dependency order, so execution follows the graph.

Your machine.
Your environment.

Work locally, with a dedicated Python environment for each project. Your analysis stays in a folder you control.

Your tools,
your agents.

Your analysis lives in ordinary Python files. Use your preferred editor or coding agent alongside Rowcall.

Drag data
onto the canvas.

Drop in CSV, TSV, or Parquet files to create Python steps that load your data. Connect them and start exploring.

How it works

  1. 01

    Start a project

    Run rowcall. The wizard guides you through naming your project, choosing a folder, and selecting Python packages.

  2. 02

    Build your analysis

    Add Python steps, connect their inputs, and inspect the results as you work.

  3. 03

    Edit and rerun

    Edit the Python files in your project folder, reopen the graph, and run it again. Share the files and requirements when you’re ready.

Rowcall in dark mode: an ecommerce Python analysis on a branching canvas, with segment revenue results in the inspector below.
An ecommerce analysis running in Rowcall. Select a step to inspect its code, inputs, and results.
Early beta

Try Rowcall.

For macOS, with Python 3.10 or newer.
Free and open source under the MIT license.

Expect bugs and changes. Keep backups of important work. Found something confusing or broken? Send feedback on GitHub ↗

1. Install from your terminal
curl -fsSL https://rowcall.io/install.sh | sh

Follow any PATH instruction printed by the installer.

2. Open the setup wizard
rowcall

Apple Silicon and Intel supported. The binary is unsigned and not notarized. Rowcall runs Python with your permissions; only open code you trust. Unsaved browser edits are lost when the app or browser closes.

Prefer to run from source? ↗ · Documentation ↗