Use case ยท July 13, 2026

Clean and prepare data offline on a Mac, from spreadsheets to datasets

Data work is exactly the kind of thing you should not upload: it often contains customer records, internal numbers or personal information. These two native Mac apps let you clean spreadsheets and prepare AI-ready datasets entirely on-device, so nothing leaves your machine. Here is how they fit together.

Product demo

See the full desktop AI workspace.

Full-frame MultiAgentOS screenshots from the current app: a built-in browser the agent drives, the Bridge chat panel, model routing, and structured results together.

  1. 1 Ask
  2. 2 Route model
  3. 3 Run tools
  4. 4 Review action
Full-frame MultiAgentOS workspace showing the navigator, the browser workspace in the centre, and the Copilot Agent Inspector with Chat, Plan, Activity, Artifacts, and Context.
Full-frame screenshot from the current MultiAgentOS app.
Browser workspace screenshot in MultiAgentOS.
Browser workspace Drive a built-in browser from the chat panel and watch every step in one window.
Visible tool runs screenshot in MultiAgentOS.
Visible tool runs Every tool call is shown with its arguments and result, so nothing happens behind your back.
Live research screenshot in MultiAgentOS.
Live research Read and act on real pages while the Bridge chat panel stays docked below.

Why offline matters for data

Spreadsheets and document sets are among the most sensitive files most people handle: customer lists, financials, HR data, personal information. Uploading them to a web tool to "clean" or "prepare" them is a data-protection risk many organizations cannot take. On-device tools remove the risk entirely, because the data never leaves your Mac.

Two stages, two tools

Data prep has two stages, and each has a focused app:

  • Tidy the tabular data. Tidyset cleans CSV, Excel and JSON with reviewable smart suggestions and previewed diffs: whitespace, inconsistent labels, missing values, fuzzy clustering to normalize near-identical labels, column profiling with histograms, and a dataset quality score. Every action becomes a reversible, reusable recipe that re-runs deterministically on next week's file.
  • Build the AI-ready dataset. Alembic takes documents and text into training-grade datasets: exact and MinHash near-duplicate removal, PII scrubbing, benchmark decontamination, semantic chunking, accurate cl100k tokenization, and eight export schemas for fine-tuning and RAG, with splits.

A private, repeatable pipeline

Together they cover the whole path from a messy export to a clean dataset, and both are built around determinism: Tidyset's recipes and Alembic's inspectable pipeline steps replay identically, so a clean-up you do once becomes a repeatable process rather than a one-off scramble. Neither uploads your data.

How to get them

On this site both are fully activated donation downloads: get Tidyset and get Alembic. Both run on-device with no subscription.

Frequently asked questions

Can I clean sensitive data without uploading it?

Yes. Tidyset and Alembic run entirely on-device, so customer data, financials and personal information never leave your Mac during cleaning or dataset preparation.

What is the difference between Tidyset and Alembic?

Tidyset cleans tabular data (CSV, Excel, JSON) with reviewable recipes. Alembic turns documents and text into training-grade datasets with deduplication, PII scrubbing, chunking and export schemas for fine-tuning and RAG.

Can I reuse the same cleaning process on new files?

Yes. Both are deterministic: Tidyset saves your steps as a reusable recipe, and Alembic's pipeline steps replay identically, so recurring data prep becomes one click.

Get the app

On this site the apps are fully activated downloads supported by a donation, with no account and no subscription. Donate and download Tidyset, or browse all nine native Mac apps.

Related reading