No account
You can use QuantFloor without creating a publisher account.
Privacy policy / Android
Your drills stay on the floor.
QuantFloor runs its practice sessions and AI coaching on your device. No account, advertising, analytics, or cloud inference is required.
At a glance
On-device
Questions, answers, scores, and coaching are not uploaded.
No tracking
No advertising SDKs, analytics tools, or data brokers.
Publisher
Who this policy covers
QuantFloor is published by Gregorius WIllson. This policy explains how the QuantFloor Android app handles information when you use it.
Local storage
Information on your device
QuantFloor keeps only the app state needed to run your practice experience:
Preferences
Tutorial completion, selected Bull or Bear coach, and reduced motion
Device only
Progress
Your high score
Device only
AI model
The downloaded Qwen3 0.6B model file
Device only
Active run
Questions, answers, attempts, and generated coaching text
In memory
Active-run information is processed in memory for the current session. The app does not upload it to the publisher or an AI service.
Network access
The model download
Before the first practice run, QuantFloor connects over HTTPS to
huggingface.co to read the public model manifest and download
the Qwen3 0.6B model from
litert-community/Qwen3-0.6B-int4. The app verifies the
downloaded file against its published checksum before installation.
Sent
Standard connection metadata, such as IP address, request time, and protocol metadata
Model host
Not sent
Questions, answers, scores, coaching text, contacts, advertising ID, and account information
Stays local
The hosting provider may process ordinary connection information under its own privacy policy. After installation, coaching inference runs locally through Google LiteRT-LM; no prompt or generated response is sent to a cloud inference service.
Third parties
Sharing and sale
QuantFloor does not sell personal information. It does not share practice activity with advertisers, analytics providers, data brokers, or social networks. The only third-party network interaction in the current release is the model download described above.
Your controls
Retention and deletion
You can delete the local coaching model from Settings. To remove all QuantFloor data, clear the app's storage in Android settings or uninstall the app. QuantFloor does not maintain a publisher account or remote practice history to delete.
Safeguards
Security
The model download uses HTTPS, is stored in Android private app storage, and is checksum-verified before use. No method of storage or transmission is perfectly secure, but QuantFloor limits its network use and does not transmit practice content.
Audience
Children
QuantFloor is intended for adults aged 18 and over and is not directed to children.
Policy page
This website
This static webpage does not set cookies and does not load advertising, analytics, or third-party scripts. The service used to host this page may process standard request metadata under its own terms.
Updates
Changes to this policy
If QuantFloor's data practices change, this policy and the Google Play Data Safety declaration will be updated before the changed version is released.
Contact
Questions?
Questions about this policy can be sent to willson2019pos@gmail.com.