1. Review the CSV
Start with Data Quality to inspect missingness, schema, risk signals, and readiness.
CSV Data Quality Checker vs AutoML
MLdeck Data Quality and MLdeck AutoML answer different questions. One reviews the CSV before decisions. The other trains and compares models.
Both workflows are browser-local and privacy-first, but neither replaces human review of what the data means.
Use Data Quality when the CSV itself is the question
Data Quality is the right first step when users want to know whether a CSV is understandable enough for reporting, cleanup, sharing, or future modeling. It checks missing values, schema clarity, risk indicators, readiness signals, and reportable review notes.
It does not train models. It does not automatically clean the file. It helps users decide what to inspect before moving forward.
Use AutoML when model comparison is the question
AutoML is for training and comparing candidate models. It helps users select targets, review metrics, inspect warnings, and export artifacts for further validation where supported.
AutoML does not remove the need to understand the data. A model can look strong while still depending on leakage, unclear schema, or weak validation evidence.
How the workflows fit together
2. Decide the next step
Document, clean, share, or continue to model training when the CSV looks suitable enough.
3. Train carefully
Use AutoML to compare models, then interpret results alongside warnings and validation scope.
4. Validate externally
Before important use, test assumptions, feature meaning, and exported artifacts outside the initial exploration.
Browser-local and privacy-first
MLdeck is designed around browser-local CSV workflows. Data Quality reviews readiness signals in the browser, and AutoML supports browser-local exploration before export and external validation steps.