Find the right ML model for your data.
AggViz is a wizard, not a black box. Point it at a dataset, and it picks the algorithm that actually fits — then shows you the visualisations that prove it. No signup.
# 891 rows · target: survived (binary)
→ recommended: GradientBoostingClassifier
→ accuracy: 0.83 · top feature: sex
→ rendered: confusion matrix, ROC, feature importance
Three steps, no setup.
From raw data to a trained model with real visualisations — in the time it takes to read this sentence.
Bring your data
Upload a CSV or pick from 19 built-in datasets — Iris, Titanic, wine, housing, digits and more. No cleanup ritual required.
The wizard reads it
AggViz profiles the shape of your data — target type, feature mix, size — and reasons about which scikit-learn family actually fits.
Model + visuals, instantly
It trains the recommended model and renders the seaborn plots that matter: distributions, correlations, confusion matrix, feature importance.
Everything you need to decide. Nothing you don't.
scikit-learn engine
Classification, regression, clustering. The same battle-tested estimators you would reach for by hand — chosen for you.
19 live datasets
Explore real datasets in one click. Great for learning, demos, or sanity-checking an approach before you wire up your own.
seaborn visualisations
Every run comes back with publication-quality plots, not just a number. See why the model behaves the way it does.
No signup, no cost
Open it and go. No account, no credit card, no rate-limit wall. Built for the moment you just need an answer.
19 datasets, ready to explore.
Click in and let the wizard work — perfect for teaching, demos, or testing an idea before you bring your own data.
Stop guessing which model to use.
Let the data tell you. Free, open, and visual from the first click.
Get started