Profile every CSV column before you clean it.
Measure completeness, uniqueness and repetition in your browser before applying a transformation. The profile updates with the preview so you can see what each cleaning step actually changes.
Find missing and repeated values
Each column shows filled and empty counts, unique values, duplicate values, and value lengths. These measures help locate weak identifiers, incomplete fields and unexpectedly repeated categories.
Compare the original with the preview
Profiles recalculate against the current cleaning plan. A shift in uniqueness or empty values becomes visible before you export, which makes unintended changes easier to catch.
Inspect locally by default
Browser-local profiling does not send the CSV to Loam or Claude. Accepted files can contain up to 100,000 rows, 500 columns and 20 MB of data.
Move to governed work when needed
Local profiling is useful for a bounded file. For durable, permissioned work on shared data, open Loam and create a source-preserving view under the existing workspace controls.