1. Start local-first and inspect your settings
A privacy-first setup is not just about turning features off. It starts by understanding what the app collects, what remains local, and what is optional. Review detector permissions and sync settings before building your baseline.
If you disable a signal category, that is fine. Just expect Sarenica to rely more heavily on the remaining signals and treat low-confidence windows as provisional.
2. Ask for coverage and reliability with every fatigue trend question
Privacy-conscious setups sometimes mean reduced signal breadth. That makes coverage and reliable minutes more important, not less. Always ask the AI to include coverage detail so you can see whether a pattern is supported.
Examples
- "Summarize my fatigue trend over the last 30 days with coverage details."
- "Compare this week vs last week and tell me if the reliable minutes are sufficient."
3. Prefer trend analysis before relationship analysis
When your data is intentionally minimal, trend summaries are usually more robust than relationship claims. First confirm that a fatigue trend exists, then ask for a relationship analysis only if coverage supports it.