Public Demo Workspace

See AutoML Services in action before uploading your own data.

Walk through real completed jobs that replay readiness checks, training progress, model comparison, saved outputs, and review pages without running new compute or charging credits.

Read-only replay Allowlisted completed jobs No credits charged

Choose a demo Create an account

What this proves

Completed-job walkthroughs

Real saved outputs from completed service workflows.

Credits charged

The public demo does not reserve, settle, or refund compute credits.

Saved evidence

Readiness, diagnostics, explainability, reproducibility, and methods appendix.

Public safety boundary

Demo pages read from allowlisted completed jobs. They do not call the training, SHAP, report-generation, prediction, billing, or job-admin APIs.

Guided scenarios

Pick the workflow you want to inspect.

Each scenario uses saved outputs from completed jobs and a replayed event stream, not live compute.

Compare all services

Cross-Sectional Classification 42 seconds

Hospital Readmission Risk

A concise classification walkthrough using a clean tabular dataset, readiness preview, model comparison, and a reviewable report summary.

Best model

CatBoost

PR AUC

0.790

1200 rows, 12 features, CSV.

Readiness Profile Service Diagnostics Explainability Status Reproducibility Manifest Methods Appendix

Start replay Completed job

Longitudinal Classification 8 minutes

Repeated-Visit Risk Stratification

A longitudinal classification walkthrough showing subject ID, time variable, readiness review, mixed-effects model comparison, and model review notes.

Best model

GPBoost Classifier

F1 Score

0.460

2391 rows, 8 features, Parquet.

Readiness Profile Service Diagnostics Explainability Status Reproducibility Manifest Methods Appendix

Start replay Completed job

Time Series Count Regression 75 seconds

Weekly Event Count Forecast

A time-series count workflow showing ordered time data, count-target checks, model comparison, and forecast-oriented review.

Best model

Count Inception

Poisson Deviance

2.760

520 rows, 6 features, XLSX.

Readiness Profile Service Diagnostics Explainability Status Reproducibility Manifest Methods Appendix

Start replay Completed job

Clustering 21 seconds

Facility Profile Segmentation

An unsupervised workflow showing no-target readiness, cluster comparison, and segment-level review.

Best model

HDBSCAN

Silhouette Score

0.420

680 rows, 10 features, Stata DTA.

Readiness Profile Service Diagnostics Explainability Status Reproducibility Manifest Methods Appendix

Start replay Completed job

After the demo

Move from replay to your own analysis.

Create an account when you are ready to upload your dataset, preview readiness, confirm the estimate, and launch a real job with saved outputs that stay with the job.

Choose service

Match data shape and study objective.

Preview readiness

Confirm roles before training controls load.

Review outputs

Inspect metrics, plots, reports, and billing.