The fastest way to see the platform working is a demo dataset from the library — no upload needed. Load one, then head to Data Synthesis.
The five-step quick start
1. Upload (or load) data. CSV with a header row. Datasets page → Upload, or load a demo.
2. Run Data Synthesis. Choose a mode (Statistical is a great first choice), set rows to generate, run.
3. Evaluate. Run evaluation on the synthetic batch and look at the Overall Quality Score and Correlation Similarity.
4. Predict. Pick a target column (e.g. conversion), train the ensemble, and read the feature-importance plots.
5. Optimize. Set objectives and constraints, run the search, and download the Pareto front.
Watching jobs
Every stage produces a job with live progress. The Runs page lists them all; a job keeps running even if you navigate away. Completed runs open a detail page with results, plots and download links.
Stuck at any step? The AI Copilot in the corner of every app page answers from this documentation, with citations. Or follow the Setting Objectives and Weights tutorial.