Bring in the messy dataset
Start from CSV, Excel, or project files already on your machine. IMD treats the raw file as the work object, not as a demo prompt.
IMD helps everyone process, clean, and deliver data with ease.
Clean table created with source snapshot preserved.
The system prepares the work. The user approves the result.
One clear path: import the file, prepare the result package, then review and export with the working context intact.
Start from CSV, Excel, or project files already on your machine. IMD treats the raw file as the work object, not as a demo prompt.
IMD detects structure, fixes inconsistent values, isolates problems, and prepares analysis-ready tables through the desktop and CLI runtime.
The user approves the result before it moves downstream into Stata, R, notebooks, or reporting workflows.
Everything on this page points back to one job: make messy working files ready for serious analysis without hiding the review process.
Missing values, inconsistent categories, duplicate keys, broken date fields, and panel-data prep are first-class workflows.
Files, snapshots, tables, and review states live in a desktop-shaped workspace instead of disappearing into a chat thread.
IMD is opinionated about automation, but the final decision remains visible and reviewable.
The app and CLI use the same account, model access, and entitlement layer.
Free resets daily. Plus and Pro reset by subscription cycle with no rollover.
The product shape keeps the workflow close to the user's machine and makes review explicit.
The goal is not to replace Stata, R, notebooks, or reporting tools. The goal is to prepare trustworthy data before those tools become useful.
Working files stay organized in a desktop-shaped workspace before results move downstream.
The desktop app, CLI, and web account use the same Supabase identity and IMD platform token layer.
AI can prepare tables and result packages, but the user remains the reviewer before export.
IMD is shaped for the unglamorous part of analysis work: making source files coherent enough that downstream methods are worth trusting.
Prepare panel, survey, and publication datasets before moving into Stata, R, or Python.
Turn messy operational files into reviewed tables with source snapshots and export-ready outputs.
Make the hidden cleanup work visible so methods, assumptions, and outputs can be reviewed.
Free users get a daily credit bucket. Plus and Pro use monthly credits that reset at the end of each subscription cycle.
Try IMD with the free model pool. Daily credits clear at reset.
For regular data cleaning work and recurring analysis preparation.
For heavier model use and larger recurring workloads.
No. Free credits clear daily, and paid credits clear when the subscription cycle renews.
No. Plus and Pro use their own monthly subscription bucket instead of stacking Free credits.
No. IMD sits upstream and prepares data so those tools can be used with more confidence.