Privacy
Plain language. No legal complexity. Updated: May 2026.
The Short Version
Tidera Health does not store your health information. Your condition, age, sex, location, medication history, and treatment details exist only in your browser during your search session. When you close the tab, they are gone.
We collect only two types of anonymized aggregate data: search events (condition searched, location, number of results) and trial feedback ratings (which trial, thumbs up/down, AI-estimated score). No personal identifiers are attached to either.
What We Do NOT Collect
- ✕Your name, email, phone number, or any contact information
- ✕Your age or date of birth
- ✕Your sex or gender identity
- ✕Your medications or medication history
- ✕Your treatment history or disease severity
- ✕Your location beyond what you voluntarily type into the search form
- ✕Your IP address
- ✕Device identifiers or browser fingerprints
- ✕Cookies for tracking or advertising
- ✕Any protected health information (PHI) as defined under HIPAA
What We Do Collect (Aggregate Only)
We collect the following anonymized data to understand how the platform is used:
Search events:
timestamp · condition (text) · location (text) · result count
Trial feedback:
timestamp · condition (text) · NCT trial ID · rating (Yes/Maybe/No) · AI fit score
Beta UX feedback:
timestamp · feedback text · would-return vote · tag · page URL
None of these events are linked to any user identity. There is no way to reconstruct who submitted any particular search or feedback from our data.
AI Processing
When you run a search, your profile (condition, age, sex, location, medications, treatment history) is sent to Anthropic's Claude API to generate plain-English trial explanations. This data is transmitted over HTTPS and used solely for the purpose of generating the response for your session.
Anthropic's data retention and privacy policies govern how they handle API inputs. Tidera Health does not instruct Anthropic to retain your data, and does not receive any personally identifiable information from Anthropic.
Explain Cache
To reduce latency on repeat searches, AI-generated trial explanations are cached in Vercel KV (Upstash Redis) for up to 12 hours. Cache keys are derived from:
- Condition slug (normalized text)
- Travel range preference
- Treatment status (has treatment history: yes/no)
- Medication class flags (biologic: yes/no; JAK inhibitor: yes/no) — booleans, not names
Medication names, age, sex, location, and any reversible patient identifiers are never part of any cache key or cached value.
Third-Party Services
- ClinicalTrials.gov (NLM/NIH): All trial data is fetched in real time. Your search query is sent to their public API. See ClinicalTrials.gov privacy policy.
- Anthropic Claude API: Used to generate plain-English explanations. See Anthropic privacy policy.
- Vercel: Hosting infrastructure. See Vercel privacy policy.
- Upstash Redis: KV cache and analytics storage. See Upstash privacy policy.
Tidera Health does not use advertising networks, social tracking pixels, or analytics services beyond the above.
Questions
If you have questions about privacy or data handling, contact us at privacy@tidera.health.
This privacy statement applies to the beta version of Tidera Health. It will be updated as the platform matures.