Turn life-story documents into an archive you can actually ask questions of — and get answers that cite the page they came from.
Most families already have the hard part done — a written history, a compiled binder, a scanned set of chapters. The problem is that nobody reads 300 pages to answer one question.
Life histories, compiled chapters, scanned typescripts. They stay exactly as they are — nothing is rewritten, and the originals are never altered.
The text is broken into passages and tagged with who is speaking and which chapter it came from, so an answer can always be traced back to a specific place.
In plain language, the way you'd ask a relative. Answers come back with their source attached, so you can go and read the passage yourself.
A chatbot that sounds confident is easy. One you'd let a relative use, about real people, is a different problem — and it's the one this is designed around.
If the source material doesn't cover something, the answer is “I don't know” — not a plausible guess. A family history that quietly invents a detail is worse than no answer at all, because nobody can tell which parts to trust.
Answers cite where they came from, so anything surprising can be checked against the original document rather than taken on faith.
People who are still living are treated differently from ancestors. Precise birth dates for the living are reduced to the year before they ever enter the searchable archive — not filtered out at the last moment, but never stored in full to begin with.
It isn't traded, either. Rather than exchange your data for free hosting, I pay for hosting and infrastructure from established providers — the more expensive choice, made deliberately.
This isn't a product you can sign up for yet. It's a real, working system — six chapters of a family history, fully indexed and answering questions — built to prove the approach before it's offered to anyone else. The live archive is private to the family whose stories are in it.
If you have a family history sitting in a binder or a folder of PDFs and you'd like it to be something people can actually ask questions of, I'd genuinely like to hear from you — including if you just want to know whether this would work for what you have.