Skip to main content
IMDIMD

Start here

Start path

Install IMD, sign in, create a project, import data, and complete your first task.

A new user should install IMD, sign in, create a project folder, import the first dataset, and then inspect the table before asking AI or creating a cleaning plan.

If you are unsure where to begin, open Get Started with IMD and follow the sections in order. Each page after that focuses on one part of the workflow.

Install IMD

Download the desktop installer from the IMD website and use the package available for your operating system. On macOS, install the app into Applications and open it from there.

If your operating system is not listed on the download page, do not assume another package will work. Use the supported installer shown by the website.

Sign in

Open IMD and choose the sign-in action. The desktop app uses the same account system as the web surface and CLI, so the account controls plan, credits, and model access.

If sign-in does not complete, try signing in from the web page first, then return to the app. On shared computers, confirm you are using the intended account before creating or opening projects.

Create project

A project is a folder on your computer. Select a folder that should contain your data work, outputs, and generated files for this project.

Use one project folder per real task or dataset family. This makes it easier to find version history, reports, and supporting assets later.

Import data

Use the import action to bring in a CSV, Excel file, or another supported tabular file. After import, the table appears in the workspace so you can inspect rows and fields.

Importing should not be treated as final cleaning. First check field types, missing values, and visible anomalies so you know what you are asking IMD or AI to help with.

First task

For a first pass, choose one small task: identify missing values, ask what a table contains, or create a basic cleaning plan for one table.

The goal is to learn the loop: open data, inspect it, ask a focused question, review the answer or plan, then decide whether to clean, export, or continue exploring.