Archive-scoped access
Private research routes require an authenticated archive membership. Cohort one is designed as one isolated deployment, database, object store, secret set, and archive—not shared multi-family tenancy.
Genealogy data is relational: one person’s upload can describe many relatives who never agreed to be part of a service. Kin Resolve treats privacy as an access, retention, and publication decision—not a footer promise.
Prelaunch disclosure updated July 15, 2026. This page describes product practice and planning; it is not an approved legal privacy notice.Imported records, research notes, cases, API responses, and unfinished hypotheses need a private place to develop. Public family history should contain only what an owner has reviewed and intentionally shared. Real-data public publishing is disabled for the first hosted cohort.
Implemented does not mean deployed or approved for private family data. Live provider configuration and launch evidence remain separate gates.
Private research routes require an authenticated archive membership. Cohort one is designed as one isolated deployment, database, object store, secret set, and archive—not shared multi-family tenancy.
Hosted DNA, external AI, binary evidence uploads, media packages, public archive access, and real-data publishing are disabled at server boundaries for cohort one.
Plain GEDCOM changes are previewed before apply, a recovery snapshot is created, and an owner can export the full archive back to GEDCOM.
Operational events use fixed names and allowlisted metadata. Record content, names, queries, credentials, response bodies, session replay, and browser recording are excluded.
Hosted accounts use single-use, expiring invitations bound to an exact archive, email, role, and approved legal-document manifest. Open signup remains disabled.
An owner can request a structured research export or record a deletion request. A deletion request is not a completed deletion; final real-pilot deletion remains an operator-reviewed whole-cell teardown.
The source includes owner-created, expiring, revocable read-only API tokens. The hosted API remains unavailable until its release, edge-limit, canary, and revocation gates pass.
The public repository, challenge, examples, and launch media use fictional Hartwell–Mercer records. Real genealogy and DNA files do not belong in source control or the beta application.
Deterministic structural and privacy checks do not need an AI provider. Although self-hosted operators can configure an OpenAI-compatible provider, the proposed hosted cohort rejects external-provider analysis at the server boundary.
If that boundary changes in a later cohort, the privacy notice must identify the provider, data sent, purpose, retention, training posture, region, and participant choice before private context leaves the deployment.
Kin Resolve does not claim GDPR, CCPA, HIPAA, or genetic-privacy compliance; multi-tenant readiness; production-grade hosted DNA handling; guaranteed backups; zero data loss; instant deletion; or an uptime SLA.
DNA, external AI, binary media, open signup, billing, shared multi-family hosting, and real-data public publishing are excluded from cohort one.
The durations below are proposed operating targets. The versioned privacy notice and participation terms control only after owner and counsel approval, publication, byte verification, and explicit participant acceptance.
The final approved privacy notice must name the providers actually configured at launch. A planned provider is not proof that a live processor relationship, region, retention rule, or contract has been approved.
Applicant email provider, Cloudflare mail routing, and the Kin Resolve beta mailbox
Contact and fixed workflow fields only; no files or family details
Vercel runtime and private object storage
Application requests and private GEDCOM artifacts under the approved product configuration
Supabase Postgres
Account, archive, research, operational, and encrypted provider-backup data
Resend
Invitation, verification, recovery, and service messages with no family-record content
A protected encrypted backup destination selected before real data
Encrypted database and both object namespaces; exact provider and expiry must be disclosed
A provider selected under the allowlisted event contract
Fixed event metadata only; no request or response bodies, record content, or replay
Participant help, export, and deletion requests will use support@kinresolve.com. Private vulnerability reports will use security@kinresolve.com. The proposed support acknowledgement target is one business day, not an SLA.
Never email family records, GEDCOM files, private screenshots, passwords, cookies, API tokens, source images, or genetic information. Arrange a separately approved private transfer only when evidence bytes are necessary.
The approved participation terms, privacy notice, and cohort boundary have not been published. No invitation should be accepted and no real family data should be uploaded until their exact versioned bytes are approved, published, verified, and presented for explicit acceptance.
A beta application consents only to beta communications; it does not accept hosted participation terms.
Read the application and cohort boundariesApply with the fixed contact and workflow fields only. Keep GEDCOM files, DNA data, source images, credentials, and private family details out of the application and email.