Kindred
A childhood vault your child can inherit
An anti-AI, zero-knowledge family archive. The answer when adding more AI is the wrong product decision and data sovereignty is the actual job to be done.
01 — The business problem
Family photos and videos default to Google, Apple and Meta clouds, where they get scanned, indexed, monetised, and quietly lost across upgrades, migrations and forgotten accounts. There is no consumer-grade alternative that a non-technical family will actually use, and no plan for what happens to the archive when the parent who curates it is no longer around.
02 — The workflow
- The parent captures the moments that define a child's story: photos, videos and notes, in one place instead of scattered across devices.
- Kindred encrypts each memory on-device before it ever leaves, then syncs to a vault in the family's own cloud. Zero-knowledge: Kindred cannot read it, scan it, or sell it.
- Trusted Guardians are added to the vault. A Heartbeat protocol periodically checks the parent is still active.
- If the parent stops responding, Guardians receive the instructions needed to release the encrypted vault to the child. The archive survives the curator.
03 — What I built
- Full product flow: landing, tiered pricing (Starter, Family, Legacy), vault creation, timeline, moments, secure backups and vault recovery.
- Auth, Guardian system, Heartbeat protocol and recovery flow shipped as working screens instead of slideware.
- PWA install banner for home-screen access. Built to be the app a non-technical parent actually opens.
- Editorial-serif identity that signals family heirloom rather than tech product. That matters for adoption in this category.
04 — Tools used
Managed backend (auth + storage) · Zero-knowledge / client-side encryption pattern · PWA · React + Vite
05 — Commercial use case
A consumer subscription with a free Starter tier plus paid Family and Legacy tiers. The same architectural pattern (client-side encryption, customer-controlled storage, no model training on user data) sells directly into regulated B2B categories such as healthcare, legal and EU public sector, where keeping AI off the data is a procurement requirement.
06 — ROI and adoption
Counter-positioned against the AI-everything default, and that is the moat.
Illustrative assumptions
- Knowing when AI is the wrong answer is itself the operator skill on display here.
- Trust and retention are the metrics this product is optimised against, ahead of hours-saved.
- The same pattern (client-side encryption, customer-owned storage, zero-knowledge guarantees) transfers directly to any B2B engagement where data sovereignty is a procurement blocker.