Sponic Garden

The AI treats sensor data as proxies but grounds optimization in human-perceived value. It correlates environmental patterns (specific lighting + music + temperature combinations) with high-value survey responses, then reinforces those states. It runs small experiments — slight variations in one variable at a time — and measures impact. Over weeks and months, the system converges on configurations that demonstrably improve member well-being, plant health, and operational efficiency.

During the early prototype phase, human experts remain in the loop — gardeners, wellness professionals, and facility operators validate AI decisions and provide training signal that pure sensors can't capture.

Right to Access (Art. 15) — Ask your agent to show all data collected about you: visit history, survey responses, booking patterns, fitness data, food preferences. Delivered in a readable dashboard, not a PDF dump.

Right to Erasure (Art. 17) — Request deletion of any or all personal data. Your agent confirms what will be deleted and executes immediately. Aggregated, anonymized data (which cannot identify you) is retained for system learning.

Right to Portability (Art. 20) — Export your complete data profile in a standard machine-readable format (JSON/CSV). Take your gardening history, wellness trends, and preferences to any other service.

Right to Restrict Processing (Art. 18) — Pause specific data uses without deleting. Example: "Keep my data but stop using it for personalized recommendations."

Right to Object (Art. 21) — Opt out of any specific processing at any time. Your agent shows you exactly what each data stream is used for and lets you toggle each one individually.

Right to Explanation (Art. 22) — Ask your agent why the AI made any recommendation. "Why did you suggest this class?" "Why was this music playing?" The system provides a human-readable explanation of the inputs and reasoning.

Tiered consent model: During onboarding, members choose their data sharing level — from "minimum" (anonymous occupancy only) to "full" (personalized recommendations, fitness integration, health insights). Every tier is fully functional; more sharing unlocks more personalization, but less sharing is never penalized.

Camera systems use on-device edge processing for occupancy counting and flow analysis. No facial recognition. No individual tracking. Raw video is never stored — only aggregate counts leave the device.

Audio sensors measure ambient energy levels (volume, density) only. Content is never recorded. Members can verify this via their privacy agent at any time.

Fitness device data is opt-in, encrypted in transit, and stored pseudonymized. Members can revoke access and trigger deletion instantly via their agent.

Surveys are anonymized at the point of submission. Individual responses are never linked to member profiles unless the member explicitly chooses identified feedback.

Data minimization: Each system collects only what it needs. Retention periods are set per data type and automatically enforced — visit logs (12 months), survey responses (aggregated after 30 days), fitness snapshots (member-controlled).

Appointed DPO: A dedicated Data Protection Officer oversees compliance, conducts regular audits, and is reachable via the member app.

Breach notification: In the unlikely event of a data breach, affected members are notified within 72 hours via their preferred channel, with plain-language explanation and remediation steps.

Vendor DPAs: Every food vendor, instructor, and third-party service signs a Data Processing Agreement. Member data never leaves the Sponic Garden platform without explicit consent.