A three-year-old in early intervention. The same child at twelve, in a school district's file. At twenty-two, aging out. At forty, looking for work. Four records, four systems, four intake forms that each begin by asking a family to explain their whole life again.
The person did not change. The database did.
And because this is the part everyone claims and nobody shows: not all of it exists yet. The identity model does — a person in our system has an id of their own, which is not their email address and does not belong to any one program. The rest is in progress. We are saying that out loud because the gap between the model and the finished thing is exactly what usually gets skipped quietly, and a family has no way to check.
What it costs a family
Families feel this as paperwork. It is not paperwork. It is what makes the difference between a provider who knows that a strategy already failed at age nine and a provider who tries it again at age fourteen because nobody wrote it down anywhere they can reach.
Three places it shows up, none of them abstract:
The transition cliff
Services are organized by age band, so the handover between bands is exactly where the record breaks. The moment a young person most needs their history to travel is the moment the system is least able to carry it. Aging out is often described as losing eligibility. Much of it is losing context.
The re-explaining tax
Ask a parent how many times they have told the same story to a new professional. The answer is a number they can recite. Every retelling costs an appointment's worth of time, and it costs something less visible: the assumption that the family are unreliable narrators of their own child's life, because nothing corroborates them.
Evidence that cannot accumulate
If a person's record restarts at each stage, nobody can see what actually worked over a decade. You get thousands of six-month snapshots and no twenty-year lines. The questions families care about most — does this help later? — are precisely the ones a fragmented record cannot answer.
What it takes to do it properly
Joining a person to themselves across time is not one clever key. It is a set of unglamorous commitments:
- An identity that outlives the program. If the identifier belongs to the service, it dies with the service. It has to belong to the person.
- Corrections that travel backwards. A name, a diagnosis, a communication preference recorded wrongly at age six must be correctable everywhere at once, or you have built a museum of old mistakes.
- Consent that is legible at every age. The person who could not consent at four can at twenty-four. A lifecycle record has to model who may see what, as that changes — and it does change.
- Refusing to guess. Two records that might be one person are not one person. A wrong merge is worse than two records, because it is invisible and it is confidently wrong. Where the evidence is not there, the honest answer is a maybe that a human resolves — not a silent join.
That last one is the part that costs money and gets skipped. It is also the one families would choose every time if anyone asked them.
Why nobody builds it
The autism world talks about the lifecycle constantly and then builds for age bands, because age bands are how services get funded and how software gets sold. The result is that the most longitudinal population imaginable — people whose needs, strengths and supports genuinely evolve over decades — is served by systems with the memory of a goldfish.
Independence is a long arc. You cannot support a long arc with short records.