Privacy Policy

Last updated: 22 June 2026

This page is an interactive educational demo operated by Kebayoran Technologies that visualizes neural-network architectures produced by automated hyperparameter optimization. This policy explains what data the demo handles.

No image uploads

The demo does not accept image uploads or use your camera. All model training is performed against a fixed set of pre-bundled, publicly available datasets (for example MNIST and CIFAR). You select a dataset and optimization settings; you do not submit any images of your own.

What we process

When you start an optimization run, the settings you choose (dataset, number of trials, epochs, and scoring weights) are sent to our self-hosted backend, which builds and trains models and returns metrics and visualizations. The resulting artifacts (metrics, plots, model summaries) are written to ephemeral server-side storage tied to the run and are not published or linked to your identity. They are not retained as a permanent record and may be cleared at any time.

Self-hosted inference

All model building, training, and inference is performed by self-hosted models on infrastructure operated by Kebayoran Technologies. No third-party inference, classification, or AI service receives your data.

No third-party tracking

This demo uses no analytics, advertising, or third-party tracking SDKs, and sets no non-essential third-party cookies. Only first-party, session-scoped state required for the application to function is used.

Contact

Questions about this policy can be directed to Kebayoran Technologies at kebayorantechnologies.com.