Dataset profiling
Summarise shape, missingness, uniqueness and field patterns before editing begins.
HypoSTAT Data Quality
CleanseIQ
Interactive previewA planned rules-led workspace for profiling datasets, reviewing anomalies and documenting each data-cleaning decision.
The product direction01 02 03 04
Make data cleaning repeatable and reviewable.
CleanseIQ is being designed around a simple principle: every change to research data should have a reason, an owner and a record. The current website preview demonstrates basic browser-side checks; it is not yet a production analysis service.
Reusable rules
Define validation checks for ranges, formats, relationships and study-specific logic.
Review queues
Separate automatic flags from human decisions so unusual records are investigated, not silently removed.
Audit trail
Track issues, decisions, versions and exports to support reproducible research.
Quality workspace
Move from “cleaned” to a defensible record of what changed.
A future workspace can connect profiling, validation rules, review notes and versioned exports.
Open the previewRows2,408
Complete94%
Review17
Workflow
Profile, test, review and export.
Profile
Inspect structure and baseline quality without changing the source file.
Test
Run transparent rules and record which checks produced each flag.
Review
Resolve exceptions with notes and role-based human oversight.
Export
Produce a versioned dataset, issue log and reproducible quality summary.
Preview limitation
The current demonstration uses small sample text in your browser and provides illustrative checks only. It must not be used for confidential data or interpreted as a validated statistical result.