About KindredKeep
The long work of remembering
Every family has a shoebox. Tapes with a grandmother's voice. A photograph nobody's quite sure about. A name on a census page three states away. The work of remembering a family — tracing the lines, saving the voices, telling the stories — is long, quiet, and often solitary.
The tools we built for that work haven't kept up. Genealogy platforms lock your tree inside their walls. Storytelling apps don't do research. Voice-archive products don't understand family. And none of them, really, feel like they belong to you.
What KindredKeep is
KindredKeep is a home for the whole thing: the tree, the records, the voices, the stories, the search. Autonomous AI agents do the digging. Your own interviews and scans become searchable memory. Conversational chat lets you ask your history questions. Every fact carries its source. And all of it runs, by default, on your own machine.
Local-first, by conviction
The cloud is useful. But your family's history shouldn't live on someone else's server by default, and it shouldn't be hostage to a subscription you might one day cancel. KindredKeep is local-first: everything works on your device, offline, forever. Cloud features (heavy research agents, voice synthesis, cross-tree collaboration) are opt-in boosts, clearly marked, run on your API keys where possible.
When you export, you get everything — GEDCOM, JSON, media, transcripts. When you delete, it's really gone.
Who this is for
Margaret, 67, a retired teacher with 20 years of genealogy research across six different tools and a box of cassette tapes. The serious hobbyist chasing a brick wall for a decade. The family storyteller who wants grandma's voice to say grandma's story, out loud, forever. The adoptee searching for a birth family. The first-generation family trying to document an oral tradition.
Built by HITS
KindredKeep is made by Howard Intelligence Technology Systems — a small studio that builds AI ingestion engines for meaningful data. KindredKeep is the one we make for ourselves.