FieldSphere Cloud – Data Ownership & AI Training Policy
FieldSphere Tech
Effective Date: YYYY-MM-DD
1. Purpose & Relationship to Privacy Policy
This Data Policy complements our Privacy Policy. While the Privacy Policy explains how we process Personal Data, this document focuses on who owns the data, how you can export it, and how your data is (or is not) used to train FieldSphere’s AI models.
2. Data Ownership
FieldSphere Tech recognises that Customers retain full and exclusive ownership of all ranch data they upload or generate through use of FieldSphere Cloud, including:
- Raw records (animal weights, pasture maps, feed inputs, photos, notes).
- Derived metrics produced for that specific ranch (feed-conversion ratio, pasture energy forecasts, health risk scores).
We access and process this data solely to deliver and improve the Service in accordance with our Terms of Service and Privacy Policy.
3. Data Export Rights
Customers may export their data at any time.
| Attribute | Details |
|---|---|
| Formats | JSON and CSV (UTF-8). |
| Scope | All raw records AND ranch-specific derived metrics; excludes anonymised global model parameters. |
| Delivery | Self-serve download button in Settings → Data & Privacy. |
| Security | Download link is a signed URL that expires after 24 hours. |
| Technical Limits | Max 5 exports per hour; individual export capped at 10 GB. For larger datasets contact support. |
4. AI Model Training & Learning Approach
FieldSphere employs a layered learning strategy:
- Personalised (Ranch-Level) Models – Trained only on that ranch’s data for tailored recommendations.
- Aggregated Global Models – Improve prediction accuracy by learning from anonymised statistics across multiple ranches.
By default, ranch data contributes to both layers.
5. Opt-Out Mechanism for Global Model Training
Customers who prefer not to have their data included in aggregated global training can opt out.
| Step | Detail |
|---|---|
| 1. Toggle | Navigate to Settings → Data & Privacy → AI Training Opt-Out and switch “Exclude my ranch from global training” to ON. |
| 2. Processing Window | Opt-out is enforced within 24 hours. |
| 3. Effect | • Data is excluded from future global model training jobs.<br>• Existing personalised models for the ranch continue to function.<br>• Past contributions remain in historical global weights but will not be retrained on. |
| 4. Confirmation | A confirmation email is sent and an entry added to the ranch’s audit log. |
6. Data Retention & Deletion
We follow the same retention rules stated in our Privacy Policy:
| Data Category | Default Retention Period |
|---|---|
| Active account data | For the life of the subscription |
| After account closure | Deleted or anonymised within 30 days |
| System logs | 12 months |
| Encrypted backups | Purged within 90 days of creation |
Upon deletion, data is removed from active systems and scheduled for destruction from backups within the stated window. Deleted data will no longer appear in exports and will not be used for future model training.
7. Change Management
We may update this Policy from time to time. Material changes will be communicated to account owners at least 30 days before they take effect. Previous versions will be archived for reference, and the “Effective Date” above will be updated.
8. Contact
Questions about this Data Policy?
Email: privacy@fieldsphere.cloud
Mail: FieldSphere Tech, [Street Address], [City], [Region], [Postal Code], [Country]
9. Definitions
- Customer – The ranch or organisation that holds a FieldSphere Cloud subscription.
- User – Any individual authorised by a Customer to use the Service.
- Personalised Model – An AI model trained solely on a single ranch’s data.
- Aggregated Global Model – An AI model trained on anonymised, combined statistics from multiple ranches.
- Signed URL – A time-limited, cryptographically signed link granting secure download access.
This policy is provided for informational purposes only and does not constitute legal advice.